notebooks/generate_stroke_examples.ipynb

{
  "nbformat": 4,
  "nbformat_minor": 0,
  "metadata": {
    "colab": {
      "name": "generate-stroke-examples.ipynb",
      "version": "0.3.2",
      "provenance": [],
      "collapsed_sections": [],
      "toc_visible": true,
      "include_colab_link": true
    },
    "kernelspec": {
      "name": "python2",
      "display_name": "Python 2"
    },
    "accelerator": "GPU"
  },
  "cells": [
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "view-in-github",
        "colab_type": "text"
      },
      "source": [
        "<a href=\"https://colab.research.google.com/github/reiinakano/diff-painter/blob/master/notebooks/generate_stroke_examples.ipynb\" target=\"_parent\"><img src=\"https://colab.research.google.com/assets/colab-badge.svg\" alt=\"Open In Colab\"/></a>"
      ]
    },
    {
      "metadata": {
        "id": "9P5f_uzAq_5g",
        "colab_type": "text"
      },
      "cell_type": "markdown",
      "source": [
        "# Install packages"
      ]
    },
    {
      "metadata": {
        "id": "LRItSKP3Kqri",
        "colab_type": "text"
      },
      "cell_type": "markdown",
      "source": [
        "## Optionally connect to Drive for saving"
      ]
    },
    {
      "metadata": {
        "id": "cl2EXKyLl06w",
        "colab_type": "code",
        "colab": {}
      },
      "cell_type": "code",
      "source": [
        "#from google.colab import drive\n",
        "#drive.mount('/drive')"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "metadata": {
        "id": "sw1EJdAmKuuj",
        "colab_type": "text"
      },
      "cell_type": "markdown",
      "source": [
        "## Install MyPaint"
      ]
    },
    {
      "metadata": {
        "id": "bAWyePtkn3av",
        "colab_type": "code",
        "colab": {}
      },
      "cell_type": "code",
      "source": [
        "# Install mypaint\n",
        "!apt-get update\n",
        "!apt-get install libjson-c-dev libgirepository1.0-dev libglib2.0-dev\n",
        "!apt-get install autotools-dev intltool gettext libtool\n",
        "!apt-get install swig python-setuptools gettext g++\n",
        "!apt-get install -y libgtk-3-dev python-gi-dev\n",
        "!apt-get install -y libpng-dev liblcms2-dev libjson-c-dev\n",
        "!apt-get install -y gir1.2-gtk-3.0 python-gi-cairo\n",
        "!apt-get install scons\n",
        "\n",
        "!wget https://github.com/mypaint/libmypaint/releases/download/v1.3.0/libmypaint-1.3.0.tar.xz\n",
        "!tar -xvf libmypaint-1.3.0.tar.xz\n",
        "!mv libmypaint-1.3.0 libmypaint\n",
        "\n",
        "!cd libmypaint && ./configure && make install\n",
        "\n",
        "!wget https://github.com/mypaint/mypaint/releases/download/v1.2.1/mypaint-1.2.1.tar.xz\n",
        "!tar -xvf mypaint-1.2.1.tar.xz\n",
        "!mv mypaint-1.2.1 mypaint\n",
        "!cd mypaint && scons && scons install\n",
        "\n",
        "!ldconfig\n",
        "\n",
        "!pip install ipdb tqdm pathlib cloudpickle future-fstrings matplotlib"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "metadata": {
        "id": "zxiqcngXKxb0",
        "colab_type": "text"
      },
      "cell_type": "markdown",
      "source": [
        "## Fetch a package for using the MyPaint environment"
      ]
    },
    {
      "metadata": {
        "id": "1ueV31ATgbQp",
        "colab_type": "code",
        "outputId": "a05cde40-0aa3-4729-865b-edf4d4db283e",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 179
        }
      },
      "cell_type": "code",
      "source": [
        "!git clone https://github.com/reiinakano/SPIRAL-tensorflow.git\n",
        "!cd SPIRAL-tensorflow && git checkout reiinakano-patch-2  #reiinakano-patches"
      ],
      "execution_count": 3,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Cloning into 'SPIRAL-tensorflow'...\n",
            "remote: Enumerating objects: 8, done.\u001b[K\n",
            "remote: Counting objects: 100% (8/8), done.\u001b[K\n",
            "remote: Compressing objects: 100% (8/8), done.\u001b[K\n",
            "remote: Total 154 (delta 3), reused 0 (delta 0), pack-reused 146\u001b[K\n",
            "Receiving objects: 100% (154/154), 1.36 MiB | 2.91 MiB/s, done.\n",
            "Resolving deltas: 100% (75/75), done.\n",
            "Branch 'reiinakano-patch-2' set up to track remote branch 'reiinakano-patch-2' from 'origin'.\n",
            "Switched to a new branch 'reiinakano-patch-2'\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "metadata": {
        "id": "a1UgoHsHK9Cp",
        "colab_type": "text"
      },
      "cell_type": "markdown",
      "source": [
        "# Imports"
      ]
    },
    {
      "metadata": {
        "id": "3ETScoLEgPJ-",
        "colab_type": "code",
        "colab": {}
      },
      "cell_type": "code",
      "source": [
        "import sys\n",
        "import os\n",
        "from pathlib import Path\n",
        "\n",
        "from PIL import Image\n",
        "import numpy as np\n",
        "\n",
        "sys.path.append('mypaint')\n",
        "sys.path.append('SPIRAL-tensorflow')\n",
        "#import envs.mnist\n",
        "\n",
        "from tqdm import tqdm\n",
        "import tensorflow as tf\n",
        "from PIL import Image, ImageDraw\n",
        "from collections import defaultdict\n",
        "\n",
        "from lib import surface, tiledsurface, brush\n",
        "from envs.mypaint_utils import *\n",
        "import matplotlib.pyplot as plt"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "metadata": {
        "id": "vRB5EmewH_x9",
        "colab_type": "code",
        "colab": {}
      },
      "cell_type": "code",
      "source": [
        ""
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "metadata": {
        "id": "bIiG_vM1K-6W",
        "colab_type": "text"
      },
      "cell_type": "markdown",
      "source": [
        "# Set up paint environment"
      ]
    },
    {
      "metadata": {
        "id": "asFw-F-ghKAU",
        "colab_type": "code",
        "colab": {}
      },
      "cell_type": "code",
      "source": [
        "class args:\n",
        "  jump=True\n",
        "  curve=True\n",
        "  screen_size=64\n",
        "  location_size=32\n",
        "  color_channel=3\n",
        "  brush_path='SPIRAL-tensorflow/assets/brushes/dry_brush.myb'\n",
        "  data_dir=Path('data')"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "metadata": {
        "id": "jT30XmogTskB",
        "colab_type": "code",
        "colab": {}
      },
      "cell_type": "code",
      "source": [
        "class PaintMode:\n",
        "  STROKES_ONLY = 0\n",
        "  JUMP_STROKES = 1\n",
        "  CONNECTED_STROKES = 2\n",
        "\n",
        "class ColorEnv():\n",
        "    head = 0.25\n",
        "    tail = 0.75\n",
        "    \n",
        "    # all 0 to 1\n",
        "    actions_to_idx = {\n",
        "        'pressure': 0,\n",
        "        'size': 1,\n",
        "        'control_x': 2,\n",
        "        'control_y': 3,\n",
        "        'end_x': 4,\n",
        "        'end_y': 5,\n",
        "        'color_r': 6,\n",
        "        'color_g': 7,\n",
        "        'color_b': 8,\n",
        "        'start_x': 9,\n",
        "        'start_y': 10,\n",
        "        'entry_pressure': 11,\n",
        "    }\n",
        "\n",
        "    def __init__(self, args, paint_mode=PaintMode.JUMP_STROKES):\n",
        "        self.args = args\n",
        "        self.paint_mode = paint_mode\n",
        "\n",
        "        # screen\n",
        "        self.screen_size = args.screen_size\n",
        "        self.height, self.width = self.screen_size, self.screen_size\n",
        "        self.observation_shape = [\n",
        "                self.height, self.width, args.color_channel]\n",
        "\n",
        "        # location\n",
        "        self.location_size = args.location_size\n",
        "        self.location_shape = [self.location_size, self.location_size]\n",
        "        \n",
        "        self.prev_x, self.prev_y, self.prev_pressure = None, None, None\n",
        "    \n",
        "    @staticmethod\n",
        "    def pretty_print_action(ac):\n",
        "        for k, v in ColorEnv.actions_to_idx.items():\n",
        "            print(k, ac[v])\n",
        "    \n",
        "    def random_action(self):\n",
        "        return np.random.uniform(size=[len(self.actions_to_idx)])\n",
        "      \n",
        "    def reset(self):\n",
        "        self.intermediate_images = []\n",
        "        self.prev_x, self.prev_y, self.prev_pressure = None, None, None\n",
        "\n",
        "        self.s = tiledsurface.Surface()\n",
        "        self.s.flood_fill(0, 0, (255, 255, 255), (0, 0, 64, 64), 0, self.s)\n",
        "        self.s.begin_atomic()\n",
        "\n",
        "        with open(self.args.brush_path) as fp:\n",
        "            self.bi = brush.BrushInfo(fp.read())\n",
        "        self.b = brush.Brush(self.bi)\n",
        "\n",
        "    def draw(self, ac, s=None, dtime=1):\n",
        "        # Just added this\n",
        "        if self.paint_mode == PaintMode.STROKES_ONLY:\n",
        "          self.s.clear()\n",
        "          self.s.flood_fill(0, 0, (255, 255, 255), (0, 0, 64, 64), 0, self.s)\n",
        "          self.s.end_atomic()\n",
        "          self.s.begin_atomic()\n",
        "        \n",
        "        if s is None:\n",
        "            s = self.s\n",
        "\n",
        "        s_x, s_y = ac[self.actions_to_idx['start_x']]*64, ac[self.actions_to_idx['start_y']]*64  \n",
        "        e_x, e_y = ac[self.actions_to_idx['end_x']]*64, ac[self.actions_to_idx['end_y']]*64\n",
        "        c_x, c_y = ac[self.actions_to_idx['control_x']]*64, ac[self.actions_to_idx['control_y']]*64\n",
        "        color = (\n",
        "            ac[self.actions_to_idx['color_r']],\n",
        "            ac[self.actions_to_idx['color_g']],\n",
        "            ac[self.actions_to_idx['color_b']],\n",
        "        )\n",
        "        pressure = ac[self.actions_to_idx['pressure']]*0.8\n",
        "        entry_pressure = ac[self.actions_to_idx['entry_pressure']]*0.8\n",
        "        size = ac[self.actions_to_idx['size']] * 2.\n",
        "        \n",
        "        if self.paint_mode == PaintMode.CONNECTED_STROKES:\n",
        "            if self.prev_x is not None:\n",
        "                s_x, s_y, entry_pressure = self.prev_x, self.prev_y, self.prev_pressure\n",
        "            self.prev_x, self.prev_y, self.prev_pressure = e_x, e_y, pressure\n",
        "\n",
        "        self.b.brushinfo.set_color_rgb(color)\n",
        "        \n",
        "        self.b.brushinfo.set_base_value('radius_logarithmic', size)\n",
        "\n",
        "        # Move brush to starting point without leaving it on the canvas.\n",
        "        self._stroke_to(s_x, s_y, 0)\n",
        "\n",
        "        self._draw(s_x, s_y, e_x, e_y, c_x, c_y, entry_pressure, pressure, size, color, dtime)\n",
        "\n",
        "    def _draw(self, s_x, s_y, e_x, e_y, c_x, c_y,\n",
        "              entry_pressure, pressure, size, color, dtime):\n",
        "\n",
        "        # if straight line or jump\n",
        "        if pressure == 0:\n",
        "            self.b.stroke_to(\n",
        "                    self.s.backend, e_x, e_y, pressure, 0, 0, dtime)\n",
        "        else:\n",
        "            self.curve(c_x, c_y, s_x, s_y, e_x, e_y, entry_pressure, pressure)\n",
        "            \n",
        "        # Relieve brush pressure for next jump\n",
        "        self._stroke_to(e_x, e_y, 0)\n",
        "\n",
        "        self.s.end_atomic()\n",
        "        self.s.begin_atomic()\n",
        "\n",
        "    # sx, sy = starting point\n",
        "    # ex, ey = end point\n",
        "    # kx, ky = curve point from last line\n",
        "    # lx, ly = last point from InteractionMode update\n",
        "    def curve(self, cx, cy, sx, sy, ex, ey, entry_pressure, pressure):\n",
        "        #entry_p, midpoint_p, junk, prange2, head, tail\n",
        "        entry_p, midpoint_p, prange1, prange2, h, t = \\\n",
        "                self._line_settings(entry_pressure, pressure)\n",
        "\n",
        "        points_in_curve = 100\n",
        "        mx, my = midpoint(sx, sy, ex, ey)\n",
        "        length, nx, ny = length_and_normal(mx, my, cx, cy)\n",
        "        cx, cy = multiply_add(mx, my, nx, ny, length*2)\n",
        "        x1, y1 = difference(sx, sy, cx, cy)\n",
        "        x2, y2 = difference(cx, cy, ex, ey)\n",
        "        head = points_in_curve * h\n",
        "        head_range = int(head)+1\n",
        "        tail = points_in_curve * t\n",
        "        tail_range = int(tail)+1\n",
        "        tail_length = points_in_curve - tail\n",
        "\n",
        "        # Beginning\n",
        "        px, py = point_on_curve_1(1, cx, cy, sx, sy, x1, y1, x2, y2)\n",
        "        length, nx, ny = length_and_normal(sx, sy, px, py)\n",
        "        bx, by = multiply_add(sx, sy, nx, ny, 0.25)\n",
        "        self._stroke_to(bx, by, entry_p)\n",
        "        pressure = abs(1/head * prange1 + entry_p)\n",
        "        self._stroke_to(px, py, pressure)\n",
        "\n",
        "        for i in xrange(2, head_range):\n",
        "            px, py = point_on_curve_1(i, cx, cy, sx, sy, x1, y1, x2, y2)\n",
        "            pressure = abs(i/head * prange1 + entry_p)\n",
        "            self._stroke_to(px, py, pressure)\n",
        "\n",
        "        # Middle\n",
        "        for i in xrange(head_range, tail_range):\n",
        "            px, py = point_on_curve_1(i, cx, cy, sx, sy, x1, y1, x2, y2)\n",
        "            self._stroke_to(px, py, midpoint_p)\n",
        "\n",
        "        # End\n",
        "        for i in xrange(tail_range, points_in_curve+1):\n",
        "            px, py = point_on_curve_1(i, cx, cy, sx, sy, x1, y1, x2, y2)\n",
        "            pressure = abs((i-tail)/tail_length * prange2 + midpoint_p)\n",
        "            self._stroke_to(px, py, pressure)\n",
        "\n",
        "        return pressure\n",
        "\n",
        "    def _stroke_to(self, x, y, pressure, duration=0.1):\n",
        "        self.b.stroke_to(\n",
        "                self.s.backend,\n",
        "                x, y,\n",
        "                pressure,\n",
        "                0.0, 0.0,\n",
        "                duration)\n",
        "        self.s.end_atomic()\n",
        "        self.s.begin_atomic()\n",
        "        self.intermediate_images.append(self.image)\n",
        "\n",
        "    def save_image(self, path=\"test.png\"):\n",
        "        Image.fromarray(self.image.astype(np.uint8).squeeze()).save(path)\n",
        "        #self.s.save_as_png(path, alpha=False)\n",
        "\n",
        "    @property\n",
        "    def image(self):\n",
        "        rect = [0, 0, self.height, self.width]\n",
        "        scanline_strips = \\\n",
        "                surface.scanline_strips_iter(self.s, rect)\n",
        "        return next(scanline_strips)\n",
        "\n",
        "    def _line_settings(self, entry_pressure, pressure):\n",
        "        p1 = entry_pressure\n",
        "        p2 = (entry_pressure + pressure) / 2\n",
        "        p3 = pressure\n",
        "        if self.head == 0.0001:\n",
        "            p1 = p2\n",
        "        prange1 = p2 - p1\n",
        "        prange2 = p3 - p2\n",
        "        return p1, p2, prange1, prange2, self.head, self.tail\n"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "metadata": {
        "id": "ukbt7gF41v-q",
        "colab_type": "code",
        "colab": {}
      },
      "cell_type": "code",
      "source": [
        ""
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "metadata": {
        "id": "EACO9tboLCPr",
        "colab_type": "text"
      },
      "cell_type": "markdown",
      "source": [
        "## Sanity check paint environment"
      ]
    },
    {
      "metadata": {
        "id": "OZ5vCnNm1q7n",
        "colab_type": "code",
        "outputId": "319d1028-66e6-4ccb-f2eb-60e85e7c20ac",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 2537
        }
      },
      "cell_type": "code",
      "source": [
        "env=ColorEnv(args, paint_mode=PaintMode.JUMP_STROKES)\n",
        "\n",
        "try:\n",
        "  os.makedirs('data/generated')\n",
        "except OSError:\n",
        "  pass\n",
        "\n",
        "for ep_idx in range(1):\n",
        "    env.reset()\n",
        "\n",
        "    for i in range(5):\n",
        "        action = env.random_action()\n",
        "        #action[env.actions_to_idx['color_r']] = 0.50\n",
        "        #action[env.actions_to_idx['color_g']] = 0.2\n",
        "        #action[env.actions_to_idx['color_b']] = 0.3\n",
        "        print(\"[Step {}] ac: {}\".format(i, action))\n",
        "        ColorEnv.pretty_print_action(action)\n",
        "        env.draw(action)\n",
        "        plt.imshow(env.image)\n",
        "        plt.show()\n",
        "        env.save_image(\"data/generated/mnist{}_{}.png\".format(ep_idx, i))"
      ],
      "execution_count": 4,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "[Step 0] ac: [0.26404815 0.96822001 0.15672115 0.82396459 0.55500448 0.25566826\n",
            " 0.02312342 0.9990442  0.50381175 0.529559   0.16601173 0.99508717]\n",
            "('end_y', 0.255668260693107)\n",
            "('end_x', 0.5550044837676973)\n",
            "('color_g', 0.9990441960515342)\n",
            "('color_b', 0.5038117495288398)\n",
            "('pressure', 0.2640481532670036)\n",
            "('entry_pressure', 0.9950871716775643)\n",
            "('color_r', 0.023123417389094736)\n",
            "('size', 0.9682200100059787)\n",
            "('control_y', 0.8239645860429625)\n",
            "('control_x', 0.1567211532692091)\n",
            "('start_x', 0.5295589957275088)\n",
            "('start_y', 0.1660117310655207)\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "iVBORw0KGgoAAAANSUhEUgAAAP4AAAD8CAYAAABXXhlaAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nQmCC\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": []
          }
        },
        {
          "output_type": "stream",
          "text": [
            "[Step 1] ac: [0.1023948  0.00936818 0.66990414 0.39761406 0.30366414 0.59681869\n",
            " 0.57223802 0.85706269 0.71255906 0.72911998 0.94880299 0.58340493]\n",
            "('end_y', 0.5968186902178001)\n",
            "('end_x', 0.30366414048940904)\n",
            "('color_g', 0.8570626923349071)\n",
            "('color_b', 0.7125590572008813)\n",
            "('pressure', 0.10239480379446841)\n",
            "('entry_pressure', 0.5834049287336872)\n",
            "('color_r', 0.5722380174290482)\n",
            "('size', 0.009368175149796154)\n",
            "('control_y', 0.39761405517650417)\n",
            "('control_x', 0.6699041443957969)\n",
            "('start_x', 0.7291199844344761)\n",
            "('start_y', 0.9488029948662642)\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "iVBORw0KGgoAAAANSUhEUgAAAP4AAAD8CAYAAABXXhlaAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nQQGJiR8EBSQmfhAUkJj4QVBA/j9bqXe92Gi4IQAAAABJRU5ErkJggg==\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": []
          }
        },
        {
          "output_type": "stream",
          "text": [
            "[Step 2] ac: [0.96587299 0.76489805 0.71957098 0.39055768 0.95974236 0.55350308\n",
            " 0.96092335 0.34436994 0.51391299 0.74046183 0.85389628 0.38754596]\n",
            "('end_y', 0.5535030804642671)\n",
            "('end_x', 0.9597423622339342)\n",
            "('color_g', 0.34436993670879856)\n",
            "('color_b', 0.5139129859386249)\n",
            "('pressure', 0.9658729929085194)\n",
            "('entry_pressure', 0.38754596201020397)\n",
            "('color_r', 0.9609233549809425)\n",
            "('size', 0.7648980535788686)\n",
            "('control_y', 0.3905576761852112)\n",
            "('control_x', 0.7195709799868061)\n",
            "('start_x', 0.7404618281675478)\n",
            "('start_y', 0.8538962786161334)\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "iVBORw0KGgoAAAANSUhEUgAAAP4AAAD8CAYAAABXXhlaAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": []
          }
        },
        {
          "output_type": "stream",
          "text": [
            "[Step 3] ac: [0.3557302  0.00308079 0.12483841 0.82961693 0.25336996 0.64261501\n",
            " 0.9151078  0.3691103  0.97120005 0.38226892 0.27619563 0.64140084]\n",
            "('end_y', 0.6426150071525691)\n",
            "('end_x', 0.2533699593771812)\n",
            "('color_g', 0.36911030044049375)\n",
            "('color_b', 0.9712000480463424)\n",
            "('pressure', 0.35573020007108636)\n",
            "('entry_pressure', 0.6414008443789228)\n",
            "('color_r', 0.9151077964708547)\n",
            "('size', 0.003080790273704337)\n",
            "('control_y', 0.8296169346434754)\n",
            "('control_x', 0.12483841406868024)\n",
            "('start_x', 0.38226892287354897)\n",
            "('start_y', 0.27619562740168646)\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "iVBORw0KGgoAAAANSUhEUgAAAP4AAAD8CAYAAABXXhlaAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": []
          }
        },
        {
          "output_type": "stream",
          "text": [
            "[Step 4] ac: [0.82820981 0.51598894 0.37106164 0.72665257 0.61908719 0.57036894\n",
            " 0.26007435 0.80208871 0.35733639 0.5359044  0.93886554 0.39163037]\n",
            "('end_y', 0.5703689352742707)\n",
            "('end_x', 0.6190871866356761)\n",
            "('color_g', 0.8020887098317845)\n",
            "('color_b', 0.3573363885998133)\n",
            "('pressure', 0.8282098107086984)\n",
            "('entry_pressure', 0.3916303740012448)\n",
            "('color_r', 0.2600743518227354)\n",
            "('size', 0.5159889409836421)\n",
            "('control_y', 0.7266525695130116)\n",
            "('control_x', 0.37106163846648565)\n",
            "('start_x', 0.5359043961075507)\n",
            "('start_y', 0.9388655449682708)\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "iVBORw0KGgoAAAANSUhEUgAAAP4AAAD8CAYAAABXXhlaAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nQmCC\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": []
          }
        }
      ]
    },
    {
      "metadata": {
        "id": "y3_6GuaK1q5M",
        "colab_type": "code",
        "outputId": "e2f3a7b9-adc0-4ae7-d3e2-3b7ea0bbac79",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 2537
        }
      },
      "cell_type": "code",
      "source": [
        "env=ColorEnv(args, paint_mode=PaintMode.STROKES_ONLY)\n",
        "\n",
        "try:\n",
        "  os.makedirs('data/generated')\n",
        "except OSError:\n",
        "  pass\n",
        "\n",
        "for ep_idx in range(1):\n",
        "    env.reset()\n",
        "\n",
        "    for i in range(5):\n",
        "        action = env.random_action()\n",
        "        #action[env.actions_to_idx['color_r']] = 0.50\n",
        "        #action[env.actions_to_idx['color_g']] = 0.2\n",
        "        #action[env.actions_to_idx['color_b']] = 0.3\n",
        "        print(\"[Step {}] ac: {}\".format(i, action))\n",
        "        ColorEnv.pretty_print_action(action)\n",
        "        env.draw(action)\n",
        "        plt.imshow(env.image)\n",
        "        plt.show()\n",
        "        env.save_image(\"data/generated/mnist{}_{}.png\".format(ep_idx, i))"
      ],
      "execution_count": 5,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "[Step 0] ac: [0.93273729 0.88879811 0.44361509 0.30820941 0.64586243 0.26765199\n",
            " 0.07082042 0.24761258 0.54161853 0.85851301 0.9840184  0.53020547]\n",
            "('end_y', 0.2676519890379069)\n",
            "('end_x', 0.6458624305453227)\n",
            "('color_g', 0.24761258061517588)\n",
            "('color_b', 0.5416185291241625)\n",
            "('pressure', 0.9327372928077287)\n",
            "('entry_pressure', 0.5302054734953655)\n",
            "('color_r', 0.0708204231802464)\n",
            "('size', 0.8887981063900303)\n",
            "('control_y', 0.30820941013104575)\n",
            "('control_x', 0.4436150901438063)\n",
            "('start_x', 0.8585130096303563)\n",
            "('start_y', 0.9840183957669343)\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "iVBORw0KGgoAAAANSUhEUgAAAP4AAAD8CAYAAABXXhlaAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nA2eFivl22TpFAAAAAElFTkSuQmCC\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": []
          }
        },
        {
          "output_type": "stream",
          "text": [
            "[Step 1] ac: [0.36515062 0.30124581 0.47204964 0.65072222 0.65151745 0.58561212\n",
            " 0.94529997 0.07012821 0.39330104 0.09341873 0.13907208 0.57115726]\n",
            "('end_y', 0.5856121174382521)\n",
            "('end_x', 0.6515174493639371)\n",
            "('color_g', 0.07012821322132035)\n",
            "('color_b', 0.39330104113802933)\n",
            "('pressure', 0.3651506189444672)\n",
            "('entry_pressure', 0.5711572595887029)\n",
            "('color_r', 0.9452999666942634)\n",
            "('size', 0.30124580694182057)\n",
            "('control_y', 0.6507222235396223)\n",
            "('control_x', 0.4720496399596482)\n",
            "('start_x', 0.09341872594776623)\n",
            "('start_y', 0.1390720766097302)\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "iVBORw0KGgoAAAANSUhEUgAAAP4AAAD8CAYAAABXXhlaAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nQgp8kQgp8EUipMAXiZACXyRC/w+zAayOzgOVTQAAAABJRU5ErkJggg==\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": []
          }
        },
        {
          "output_type": "stream",
          "text": [
            "[Step 2] ac: [0.09476984 0.23737978 0.45335428 0.42428263 0.4455534  0.65612943\n",
            " 0.19244729 0.01741144 0.76810654 0.01604109 0.57286472 0.10633994]\n",
            "('end_y', 0.6561294325314068)\n",
            "('end_x', 0.44555340086482953)\n",
            "('color_g', 0.017411440663484545)\n",
            "('color_b', 0.768106540693279)\n",
            "('pressure', 0.09476983659940319)\n",
            "('entry_pressure', 0.10633994436518801)\n",
            "('color_r', 0.19244729181189735)\n",
            "('size', 0.23737978067958754)\n",
            "('control_y', 0.42428262690577456)\n",
            "('control_x', 0.45335428471462236)\n",
            "('start_x', 0.016041085908013786)\n",
            "('start_y', 0.5728647197065228)\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "iVBORw0KGgoAAAANSUhEUgAAAP4AAAD8CAYAAABXXhlaAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nfLMEOfhmCfp/GU+IMZ4AsPEAAAAASUVORK5CYII=\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": []
          }
        },
        {
          "output_type": "stream",
          "text": [
            "[Step 3] ac: [0.56323853 0.60769168 0.24782252 0.95826642 0.35986308 0.21879381\n",
            " 0.3866676  0.53685841 0.00205911 0.71798041 0.34658607 0.87065781]\n",
            "('end_y', 0.21879380981931118)\n",
            "('end_x', 0.3598630810590875)\n",
            "('color_g', 0.5368584072771373)\n",
            "('color_b', 0.002059114656629779)\n",
            "('pressure', 0.5632385291365636)\n",
            "('entry_pressure', 0.8706578070325454)\n",
            "('color_r', 0.3866676046596915)\n",
            "('size', 0.6076916789220893)\n",
            "('control_y', 0.9582664209477781)\n",
            "('control_x', 0.24782251608811356)\n",
            "('start_x', 0.7179804087376247)\n",
            "('start_y', 0.34658606848380813)\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "iVBORw0KGgoAAAANSUhEUgAAAP4AAAD8CAYAAABXXhlaAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nFgAAAABJRU5ErkJggg==\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": []
          }
        },
        {
          "output_type": "stream",
          "text": [
            "[Step 4] ac: [0.46326046 0.68938223 0.81022043 0.31715687 0.56169805 0.53509667\n",
            " 0.57993134 0.47325214 0.32227619 0.00099051 0.53090457 0.7133904 ]\n",
            "('end_y', 0.5350966703563642)\n",
            "('end_x', 0.5616980510440263)\n",
            "('color_g', 0.4732521440564966)\n",
            "('color_b', 0.3222761863805066)\n",
            "('pressure', 0.46326046078452465)\n",
            "('entry_pressure', 0.7133903961725964)\n",
            "('color_r', 0.5799313384080698)\n",
            "('size', 0.6893822342459711)\n",
            "('control_y', 0.3171568685266385)\n",
            "('control_x', 0.8102204263584482)\n",
            "('start_x', 0.0009905088440437249)\n",
            "('start_y', 0.5309045650884464)\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "iVBORw0KGgoAAAANSUhEUgAAAP4AAAD8CAYAAABXXhlaAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": []
          }
        }
      ]
    },
    {
      "metadata": {
        "id": "Hd4jqVXg1q3L",
        "colab_type": "code",
        "outputId": "290cb5a8-1581-493c-8d79-173ee2187b81",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 2537
        }
      },
      "cell_type": "code",
      "source": [
        "env=ColorEnv(args, paint_mode=PaintMode.CONNECTED_STROKES)\n",
        "\n",
        "try:\n",
        "  os.makedirs('data/generated')\n",
        "except OSError:\n",
        "  pass\n",
        "\n",
        "for ep_idx in range(1):\n",
        "    env.reset()\n",
        "\n",
        "    for i in range(5):\n",
        "        action = env.random_action()\n",
        "        #action[env.actions_to_idx['color_r']] = 0.50\n",
        "        #action[env.actions_to_idx['color_g']] = 0.2\n",
        "        #action[env.actions_to_idx['color_b']] = 0.3\n",
        "        action[env.actions_to_idx['entry_pressure']] = 0.\n",
        "        print(\"[Step {}] ac: {}\".format(i, action))\n",
        "        ColorEnv.pretty_print_action(action)\n",
        "        env.draw(action)\n",
        "        plt.imshow(env.image)\n",
        "        plt.show()\n",
        "        env.save_image(\"data/generated/mnist{}_{}.png\".format(ep_idx, i))"
      ],
      "execution_count": 6,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "[Step 0] ac: [0.32542663 0.37338011 0.36733832 0.48550354 0.23065611 0.23097348\n",
            " 0.45006508 0.21099824 0.16291906 0.24169058 0.72456516 0.        ]\n",
            "('end_y', 0.230973482733566)\n",
            "('end_x', 0.23065611341710412)\n",
            "('color_g', 0.21099823920283978)\n",
            "('color_b', 0.16291906446897064)\n",
            "('pressure', 0.325426627643262)\n",
            "('entry_pressure', 0.0)\n",
            "('color_r', 0.45006507593491396)\n",
            "('size', 0.3733801136177013)\n",
            "('control_y', 0.48550354234559234)\n",
            "('control_x', 0.36733832463080496)\n",
            "('start_x', 0.24169057786216064)\n",
            "('start_y', 0.7245651603918453)\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "iVBORw0KGgoAAAANSUhEUgAAAP4AAAD8CAYAAABXXhlaAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nb5ag/wPMgbHGJzFd4gAAAABJRU5ErkJggg==\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": []
          }
        },
        {
          "output_type": "stream",
          "text": [
            "[Step 1] ac: [0.03044385 0.43777396 0.35321483 0.34914334 0.26010034 0.87082733\n",
            " 0.23282752 0.05244238 0.01913243 0.34816328 0.41009838 0.        ]\n",
            "('end_y', 0.8708273330860063)\n",
            "('end_x', 0.26010034026881035)\n",
            "('color_g', 0.05244237768434523)\n",
            "('color_b', 0.0191324263207161)\n",
            "('pressure', 0.03044384524715349)\n",
            "('entry_pressure', 0.0)\n",
            "('color_r', 0.23282752039353372)\n",
            "('size', 0.43777396232244514)\n",
            "('control_y', 0.3491433391822223)\n",
            "('control_x', 0.3532148303589294)\n",
            "('start_x', 0.3481632764775672)\n",
            "('start_y', 0.4100983775675553)\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "iVBORw0KGgoAAAANSUhEUgAAAP4AAAD8CAYAAABXXhlaAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nkiAFvkiCFPgiCfo/BqKgpfQ7/XgAAAAASUVORK5CYII=\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": []
          }
        },
        {
          "output_type": "stream",
          "text": [
            "[Step 2] ac: [0.00355761 0.99963391 0.76194057 0.83948911 0.66236439 0.80897455\n",
            " 0.32711967 0.9174523  0.16179189 0.1977505  0.13337123 0.        ]\n",
            "('end_y', 0.8089745475184195)\n",
            "('end_x', 0.6623643930583905)\n",
            "('color_g', 0.9174522979697065)\n",
            "('color_b', 0.16179188837154068)\n",
            "('pressure', 0.0035576058829480672)\n",
            "('entry_pressure', 0.0)\n",
            "('color_r', 0.3271196731498318)\n",
            "('size', 0.999633912948483)\n",
            "('control_y', 0.8394891101499349)\n",
            "('control_x', 0.7619405736245075)\n",
            "('start_x', 0.1977504994387269)\n",
            "('start_y', 0.13337123415109253)\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "iVBORw0KGgoAAAANSUhEUgAAAP4AAAD8CAYAAABXXhlaAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nJ11Q2qx7AAAAAElFTkSuQmCC\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": []
          }
        },
        {
          "output_type": "stream",
          "text": [
            "[Step 3] ac: [0.20891175 0.92287791 0.72349815 0.296454   0.92919243 0.24524658\n",
            " 0.22189355 0.04993547 0.33627227 0.45439776 0.69748406 0.        ]\n",
            "('end_y', 0.24524658330318427)\n",
            "('end_x', 0.9291924255303122)\n",
            "('color_g', 0.04993547376523011)\n",
            "('color_b', 0.3362722725734334)\n",
            "('pressure', 0.20891174905536225)\n",
            "('entry_pressure', 0.0)\n",
            "('color_r', 0.2218935537794937)\n",
            "('size', 0.922877905665151)\n",
            "('control_y', 0.2964539985294652)\n",
            "('control_x', 0.7234981475387975)\n",
            "('start_x', 0.4543977553412989)\n",
            "('start_y', 0.6974840583596295)\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "iVBORw0KGgoAAAANSUhEUgAAAP4AAAD8CAYAAABXXhlaAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\ncOITEiGc+IREyP8Hu8hgBI8XqYkAAAAASUVORK5CYII=\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": []
          }
        },
        {
          "output_type": "stream",
          "text": [
            "[Step 4] ac: [0.83328842 0.98344846 0.33902685 0.89123158 0.52925477 0.91303524\n",
            " 0.24966657 0.2142516  0.83175996 0.96174141 0.73542546 0.        ]\n",
            "('end_y', 0.9130352411236383)\n",
            "('end_x', 0.5292547683514356)\n",
            "('color_g', 0.2142515993495543)\n",
            "('color_b', 0.8317599640955169)\n",
            "('pressure', 0.8332884207039201)\n",
            "('entry_pressure', 0.0)\n",
            "('color_r', 0.24966657442185525)\n",
            "('size', 0.9834484610593238)\n",
            "('control_y', 0.891231583965259)\n",
            "('control_x', 0.3390268529310515)\n",
            "('start_x', 0.9617414052565931)\n",
            "('start_y', 0.7354254600146377)\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "iVBORw0KGgoAAAANSUhEUgAAAP4AAAD8CAYAAABXXhlaAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\n+MB3nBTiA99xUogPfMdJIT7wHSeF/H8/VK55qrUdMQAAAABJRU5ErkJggg==\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": []
          }
        }
      ]
    },
    {
      "metadata": {
        "id": "oceF6vnA1q0D",
        "colab_type": "code",
        "colab": {}
      },
      "cell_type": "code",
      "source": [
        ""
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "metadata": {
        "id": "QmmnGkLU1qv8",
        "colab_type": "code",
        "colab": {}
      },
      "cell_type": "code",
      "source": [
        ""
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "metadata": {
        "id": "FG8PvckglJFO",
        "colab_type": "code",
        "colab": {}
      },
      "cell_type": "code",
      "source": [
        ""
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "metadata": {
        "id": "MidgiI1rLe_h",
        "colab_type": "text"
      },
      "cell_type": "markdown",
      "source": [
        "# Generate action to brushstroke image examples\n",
        "\n",
        "This takes a long time. You might want to adjust the number of shards or change the Colab runtime to local or save directly to Google Drive to avoid losing progress.\n",
        "\n",
        "Alternatively, just skip this step and download the strokes I generated."
      ]
    },
    {
      "metadata": {
        "id": "3bSfbLaLjdHv",
        "colab_type": "code",
        "colab": {}
      },
      "cell_type": "code",
      "source": [
        "env=ColorEnv(args, paint_mode=PaintMode.STROKES_ONLY)\n",
        "\n",
        "try:\n",
        "  os.makedirs('data')\n",
        "except OSError:\n",
        "  pass\n",
        "\n",
        "NUM_SHARDS = 80\n",
        "NUM_STROKES_PER_SHARD = 100000\n",
        "\n",
        "SHARD_NUM_OFFSET = 0\n",
        "\n",
        "for i in range(SHARD_NUM_OFFSET, SHARD_NUM_OFFSET+NUM_SHARDS):\n",
        "  print('shard', i)\n",
        "  actions = []\n",
        "  strokes = []\n",
        "  for idx in range(NUM_STROKES_PER_SHARD):\n",
        "    env.reset()\n",
        "    if idx % 2000 == 0: print(idx)\n",
        "    \n",
        "    action = env.random_action()\n",
        "    actions.append(action)\n",
        "    env.draw(action)\n",
        "    strokes.append(env.image[:, :, :3])\n",
        "  actions = np.array(actions, dtype=np.float)\n",
        "  strokes = np.array(strokes, dtype=np.uint8)\n",
        "  np.savez_compressed(\"data/episodes_{}.npz\".format(i), actions=actions, strokes=strokes)\n"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "metadata": {
        "id": "EwhdhVkdM7ZV",
        "colab_type": "text"
      },
      "cell_type": "markdown",
      "source": [
        "# Test out the generated files"
      ]
    },
    {
      "metadata": {
        "id": "2z3CXRd1TRma",
        "colab_type": "code",
        "colab": {}
      },
      "cell_type": "code",
      "source": [
        "loaded = np.load('data/episodes_0.npz')"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "metadata": {
        "id": "-3L2cE3djdaQ",
        "colab_type": "code",
        "outputId": "4c4f8095-812e-40b9-9d2e-fb9028db2696",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 53
        }
      },
      "cell_type": "code",
      "source": [
        "print loaded['actions'].shape\n",
        "print loaded['strokes'].shape\n"
      ],
      "execution_count": 9,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "(100000, 12)\n",
            "(100000, 64, 64, 3)\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "metadata": {
        "id": "uIWBkyd_jdV0",
        "colab_type": "code",
        "outputId": "b3f7e791-f05c-4ebe-87e3-8e4c3ad8c778",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 122
        }
      },
      "cell_type": "code",
      "source": [
        "import matplotlib.pyplot as plt\n",
        "\n",
        "w=args.screen_size\n",
        "h=args.screen_size\n",
        "fig=plt.figure(figsize=(30, 10))\n",
        "cols = 10\n",
        "\n",
        "smaller_arr = loaded['strokes'][:cols]\n",
        "\n",
        "for col in range(cols):\n",
        "    img = smaller_arr[col][:, :, :3]\n",
        "    #print(img.shape)\n",
        "    fig.add_subplot(1, 20, col+1)\n",
        "    plt.grid(False)\n",
        "    plt.imshow(img)\n",
        "plt.show()"
      ],
      "execution_count": 10,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "image/png": "iVBORw0KGgoAAAANSUhEUgAAA2MAAABpCAYAAAC3QvvyAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nJHII/P/yRfqRqtKl0QAAAABJRU5ErkJggg==\n",
            "text/plain": [
              "<Figure size 2160x720 with 10 Axes>"
            ]
          },
          "metadata": {
            "tags": []
          }
        }
      ]
    },
    {
      "metadata": {
        "id": "l4NbuBaiF3eo",
        "colab_type": "code",
        "colab": {}
      },
      "cell_type": "code",
      "source": [
        "for ac in loaded['actions'][:cols]:\n",
        "  print('___')\n",
        "  ColorEnv.pretty_print_action(ac)"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "metadata": {
        "id": "s7vyPqLbjdQm",
        "colab_type": "code",
        "colab": {}
      },
      "cell_type": "code",
      "source": [
        ""
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "metadata": {
        "id": "shl-Vn7tjdNd",
        "colab_type": "code",
        "colab": {}
      },
      "cell_type": "code",
      "source": [
        "\"\"\"\n",
        "\n",
        "env=ColorEnv(args, paint_mode=PaintMode.JUMP_STROKES)\n",
        "#env2 = ColorEnv(args, paint_mode=PaintMode.JUMP_STROKES)\n",
        "\n",
        "try:\n",
        "  os.makedirs('data')\n",
        "except OSError:\n",
        "  pass\n",
        "\n",
        "NUM_STROKES = 200000\n",
        "NUM_SHARDS = 1\n",
        "NUM_STROKES_PER_SHARD = NUM_STROKES/NUM_SHARDS\n",
        "\n",
        "for i in range(NUM_SHARDS):\n",
        "  env.reset()\n",
        "  #env2.reset()\n",
        "  actions = []\n",
        "  strokes = []\n",
        "  #full_images = []\n",
        "  for idx in range(NUM_STROKES_PER_SHARD):\n",
        "    if idx % 2000 == 0: print(idx)\n",
        "    \n",
        "    action = env.random_action()\n",
        "    actions.append(action)\n",
        "    env.draw(action)\n",
        "    #env2.draw(action)\n",
        "    strokes.append(env.image[:, :, :3])\n",
        "    #full_images.append(env2.image[:, :, :3])\n",
        "  actions = np.array(actions, dtype=np.float)\n",
        "  strokes = np.array(strokes, dtype=np.uint8)\n",
        "  #full_images = np.array(full_images, dtype=np.uint8)\n",
        "  np.savez_compressed(\"data/full_strokes_{}.npz\".format(i), actions=actions, strokes=strokes)\n",
        "\"\"\""
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "metadata": {
        "id": "_D9ezTtHjdFq",
        "colab_type": "code",
        "colab": {}
      },
      "cell_type": "code",
      "source": [
        ""
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "metadata": {
        "id": "GIq_7RrCjdCW",
        "colab_type": "code",
        "colab": {}
      },
      "cell_type": "code",
      "source": [
        ""
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "metadata": {
        "id": "9FDftk_6jdAk",
        "colab_type": "code",
        "colab": {}
      },
      "cell_type": "code",
      "source": [
        ""
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "metadata": {
        "id": "Et17CN9Xjc-1",
        "colab_type": "code",
        "colab": {}
      },
      "cell_type": "code",
      "source": [
        ""
      ],
      "execution_count": 0,
      "outputs": []
    }
  ]
}