legend.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Legend"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 576x576 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# -----------------------------------------------------------------------------\n",
    "# Matplotlib cheat book\n",
    "# Released under the BSD License\n",
    "# -----------------------------------------------------------------------------\n",
    "import numpy as np\n",
    "import matplotlib as mpl\n",
    "import matplotlib.pyplot as plt\n",
    "\n",
    "\n",
    "fig = plt.figure(figsize=(8,8))\n",
    "ax = fig.add_axes([0.15,0.15,.7,.7], frameon=True, aspect=1,\n",
    "                  xticks=[], yticks=[])\n",
    "\n",
    "def text(x, y, _text):\n",
    "    color= \"C1\"\n",
    "    if not 0 < x < 1 or not 0 < y < 1:  color = \"C0\"\n",
    "    size = 0.15\n",
    "    ax.text(x, y, _text, color=\"white\", #bbox={\"color\": \"C1\"},\n",
    "            size=\"xx-large\", weight=\"bold\", ha=\"center\", va=\"center\")\n",
    "    rect = plt.Rectangle((x-size/2, y-size/2), size, size, facecolor=color,\n",
    "                         zorder=-10, clip_on=False)\n",
    "    ax.add_patch(rect)\n",
    "\n",
    "def point(x, y):\n",
    "    ax.scatter([x], [y], facecolor=\"C0\", edgecolor=\"white\",\n",
    "               zorder=10, clip_on=False)\n",
    "\n",
    "d = .1\n",
    "e = .15/2\n",
    "\n",
    "text(  d,   d, \"1\"), text( 0.5,   d, \"2\"), text(1-d,   d, \"3\")\n",
    "text(  d, 0.5, \"4\"), text( 0.5, 0.5, \"5\"), text(1-d, 0.5, \"6\")\n",
    "text(  d, 1-d, \"7\"), text( 0.5, 1-d, \"8\"), text(1-d, 1-d, \"9\")\n",
    "\n",
    "text( -d, 1-d, \"A\"), text( -d, 0.5, \"B\"), text(  -d,   d, \"C\")\n",
    "point(-d+e, 1-d+e), point(-d+e, 0.5), point(-d+e, d-e),\n",
    "\n",
    "text(  d,  -d, \"D\"), text(0.5,  -d, \"E\"), text( 1-d,  -d, \"F\")\n",
    "point(d-e, -d+e), point(0.5, -d+e), point(1-d+e, -d+e),\n",
    "\n",
    "text(1+d,   d, \"G\"), text(1+d, 0.5, \"H\"), text( 1+d, 1-d, \"I\")\n",
    "point(1+d-e, d-e), point(1+d-e, .5), point(1+d-e, 1-d+e),\n",
    "\n",
    "text(1-d, 1+d, \"J\"), text(0.5, 1+d, \"K\"), text(   d, 1+d, \"L\")\n",
    "point(1-d+e, 1+d-e), point(0.5, 1+d-e), point(d-e, 1+d-e),\n",
    "\n",
    "plt.xlim(0,1), plt.ylim(0,1)\n",
    "\n",
    "# plt.savefig(\"../figures/legend-placement.pdf\")\n",
    "plt.show()"
   ]
  },
  {
   "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": "3.6.5"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}