{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "%reload_ext autoreload\n",
    "%autoreload 2\n",
    "\n",
    "import requests\n",
    "from io import BytesIO\n",
    "import numpy as np\n",
    "from PIL import Image\n",
    "import matplotlib.pyplot as plt\n",
    "import matplotlib.colors as mcolors"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Hues on Color Wheel\n",
    "\n",
    "On the color wheel, hues form a circle — not a straight line. Reds live near both 0° and 360° (or near 0 and 1 in normalized format). \n",
    "\n",
    "Naive algorithm may mistakenly think there are two separate groups of reds instead of the true one."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "url = \"https://github.com/timpyrkov/circleclust/blob/master/tests/image.jpg?raw=true\"\n",
    "\n",
    "# Read image from url (image.jpg from github repository)\n",
    "response = requests.get(url)\n",
    "img = Image.open(BytesIO(response.content))\n",
    "\n",
    "# Show image\n",
    "plt.imshow(img)\n",
    "plt.axis('off')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Convert to RGB array\n",
    "if img.mode != 'RGB':\n",
    "    img = img.convert('RGB')\n",
    "rgb = np.array(img)  / 255\n",
    "\n",
    "# Convert HSV and extract Hue (H) coordinates on the color wheel\n",
    "hsv = mcolors.rgb_to_hsv(rgb)\n",
    "hue = hsv[:, :, 0].flatten()\n",
    "\n",
    "# Show histogram of hue values in range [0,1)\n",
    "plt.figure(figsize=(10, 5))\n",
    "plt.title('Histogram of pixel hues', fontsize=14)\n",
    "plt.hist(hue, bins=np.linspace(0, 1, 50), width=.017, alpha=.3)\n",
    "\n",
    "plt.xticks([0.0, 0.17, 0.33, 0.5, 0.67, 0.83, 1.0], \n",
    "           [\"Reds\", \"Yellows\", \"Greens\", \"Cyans\", \"Blues\", \"Magentas\", \"Reds\"])\n",
    "\n",
    "plt.show()\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Find Dominant Hues\n",
    " \n",
    " Detect dominant pixel hues (\"Reds\" and \"Magentas\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "from circleclust import CircleClust\n",
    "\n",
    "# Initialize and run CircleClust.fit() to find pixel group centroids\n",
    "cc = CircleClust()\n",
    "cc.fit(hue, period=1) # Important: provide correct range of data values period!\n",
    "\n",
    "# Print detected centroids\n",
    "cc.get_centroids()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Show detected centroids\n",
    "cc.show_centroids()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Hue Clusters Size\n",
    "\n",
    "Note that the \"Reds\" cluster occupies the area of wrap near Hue = 0. Both pixels close to 0  and close to 1 are attributed to the same single \"Reds\" cluster."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Run CircleClust.predict() to assign cluster labels\n",
    "clust_ids = cc.predict(hue, width_scale=1.0)\n",
    "\n",
    "# Print number of each cluster labels (-1 stands for outliers)\n",
    "np.unique(clust_ids, return_counts=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Show histogram of each pixel hue cluster\n",
    "plt.figure(figsize=(10, 5))\n",
    "plt.title('Histogram of pixel hues per cluster', fontsize=14)\n",
    "\n",
    "bins = np.linspace(0, 1, 50)\n",
    "color = [\"grey\", \"red\", \"magenta\"]\n",
    "name = [\"Outliers\", \"Reds\", \"Magentas\"]\n",
    "clist, ccount = np.unique(clust_ids, return_counts=True)\n",
    "for i, clust_id in enumerate(clist):\n",
    "    mask = clust_ids == clust_id\n",
    "    label = f\"{name[i]}: {ccount[i]:_}\".replace(\"_\", \"'\")\n",
    "    plt.hist(hue[mask], bins=bins, width=.017, alpha=.3, color=color[i], label=label)\n",
    "\n",
    "plt.xticks([0.0, 0.17, 0.33, 0.5, 0.67, 0.83, 1.0], \n",
    "           [\"Reds\", \"Yellows\", \"Greens\", \"Cyans\", \"Blues\", \"Magentas\", \"Reds\"])\n",
    "\n",
    "plt.legend()\n",
    "plt.show()\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3 (ipykernel)",
   "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": "3.12.11"
  },
  "varInspector": {
   "cols": {
    "lenName": 16,
    "lenType": 16,
    "lenVar": 40
   },
   "kernels_config": {
    "python": {
     "delete_cmd_postfix": "",
     "delete_cmd_prefix": "del ",
     "library": "var_list.py",
     "varRefreshCmd": "print(var_dic_list())"
    },
    "r": {
     "delete_cmd_postfix": ") ",
     "delete_cmd_prefix": "rm(",
     "library": "var_list.r",
     "varRefreshCmd": "cat(var_dic_list()) "
    }
   },
   "types_to_exclude": [
    "module",
    "function",
    "builtin_function_or_method",
    "instance",
    "_Feature"
   ],
   "window_display": false
  }
 },
 "nbformat": 4,
 "nbformat_minor": 1
}