IG_clustering.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "id": "65233dec-cda0-4612-bc67-dbbc78f38c27",
   "metadata": {},
   "source": [
    "# Домашнее задание к лекции «Алгоритмы кластеризации»"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "0036e4c7-5b94-4eef-8d30-7e3132340d24",
   "metadata": {},
   "source": [
    "## **Цель:**\n",
    "Изучить применение моделей кластеризации"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "6fd94606-0dd6-4dff-b2a7-377be76d6890",
   "metadata": {},
   "source": [
    "### **Описание задания:**\n",
    "В домашнем задании нужно сократить число цветов в палитре изображения. Картинку для выполнения работы можно выбрать любую, главное условие – наличие на ней разных цветов, для того, чтобы результат работы моделей был заметен.\n",
    "Для выполнения работы необходимо выделить кластеры в пространстве RGB, объекты соответствуют пикселям изображения. После выделения кластеров все пиксели, отнесенные в один кластер, заполняются одним цветом. Цвет – центроид соответствующего кластера."
   ]
  },
  {
   "cell_type": "markdown",
   "id": "b20d301f-2996-48d9-8ce0-7a5e944f8586",
   "metadata": {},
   "source": [
    "### **Этапы работы:**"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "da3e8055-be04-4dc8-9d69-79282fad8143",
   "metadata": {},
   "source": [
    "#### 1. Получите данные и загрузите их в рабочую среду (Jupyter Notebook или другую).  \n",
    "a) Для отображения картинки воспользуйтесь вспомогательной функцией:  \n",
    "import matplotlib.pyplot as plt  \n",
    "import cv2  \n",
    "import skimage  \n",
    "def draw_picture(image, title, bgr=False):  \n",
    "b, g, r = cv2.split(image) # по умолчанию cv2 почему-то отдает цвета в порядке BGR вместо RGB  \n",
    "new_image = cv2.merge([r, g, b])  \n",
    "plt.figure(figsize=(7, 5))  \n",
    "plt.axis(‘off’)  \n",
    "plt.imshow(new_image)  \n",
    "plt.title(title)  \n",
    "plt.show()  \n",
    "b) Для загрузки картинки можно использовать код:  \n",
    "img = cv.imread(path)  "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "id": "3c29fdd2-fbdf-4631-b252-e16a6df1e51d",
   "metadata": {},
   "outputs": [],
   "source": [
    "import cv2\n",
    "import skimage\n",
    "from skimage.metrics import structural_similarity as ssim"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "id": "a7f19a4f-34dd-42b9-b26f-6974d18a960d",
   "metadata": {},
   "outputs": [],
   "source": [
    "import pandas as pd\n",
    "import numpy as np"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "id": "0816984e-3c1b-44ad-86eb-d6f365fbb8b0",
   "metadata": {},
   "outputs": [],
   "source": [
    "from matplotlib import pyplot as plt\n",
    "%matplotlib inline\n",
    "import seaborn as sns\n",
    "\n",
    "#%config InlineBackend.figure_format = 'retina'\n",
    "#from pylab import rcParams\n",
    "#rcParams['figure.figsize'] = (12, 9)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "id": "fd959bb8-b89a-4963-a2d8-cacda08da397",
   "metadata": {
    "scrolled": true
   },
   "outputs": [],
   "source": [
    "def draw_picture(image, title, bgr=False):\n",
    "    b, g, r = cv2.split(image) # by default cv2 works with BGR order instead of RGB\n",
    "    new_img = cv2.merge([r, g, b])\n",
    "    plt.figure(figsize=(7, 5))\n",
    "    plt.axis('off')\n",
    "    plt.imshow(new_img)\n",
    "    plt.title(title)\n",
    "    plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "id": "e5f4e650-cd04-4feb-aa34-7d98d87a7032",
   "metadata": {},
   "outputs": [],
   "source": [
    "image_org = cv2.imread('C://Users//grebe//Pictures//picture.webp')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "id": "05151c87-45f0-40f7-8e19-a1e1d2f1a520",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(743, 940, 3)"
      ]
     },
     "execution_count": 6,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "image_org.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "id": "0d37c142-ea7c-410d-8679-566e86f5b4db",
   "metadata": {},
   "outputs": [],
   "source": [
    "img = cv2.resize(image_org, (128, 128)) # reduced size for faster clustering"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "id": "8c92db4a-3a97-42d8-b5b7-4d14d3b8bf4f",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 700x500 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "draw_picture(img,'Original')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "id": "f82f65b1-39b5-45b6-b17e-da1e6da0cb65",
   "metadata": {},
   "outputs": [],
   "source": [
    "X_train = pd.DataFrame(img.reshape((-1,3)), columns = ['B','G','R']) # transformed to dataframe"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "id": "ad6e70d5-e45d-4e49-84b7-2ca6249221fe",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>B</th>\n",
       "      <th>G</th>\n",
       "      <th>R</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>253</td>\n",
       "      <td>255</td>\n",
       "      <td>253</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>253</td>\n",
       "      <td>255</td>\n",
       "      <td>253</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>253</td>\n",
       "      <td>255</td>\n",
       "      <td>255</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>251</td>\n",
       "      <td>255</td>\n",
       "      <td>255</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>251</td>\n",
       "      <td>255</td>\n",
       "      <td>255</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "     B    G    R\n",
       "0  253  255  253\n",
       "1  253  255  253\n",
       "2  253  255  255\n",
       "3  251  255  255\n",
       "4  251  255  255"
      ]
     },
     "execution_count": 10,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "X_train.head()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "e006f1fb-06ef-4527-9056-2971a262ed5d",
   "metadata": {},
   "source": [
    "#### 2. Реализуйте три модели кластеризации:\n",
    "a) [KMeans](https://scikit-learn.org/stable/modules/generated/sklearn.cluster.KMeans.html). Рассмотрите число кластеров K = 2, 5, 10, 20.  \n",
    "b) [DBSCAN](https://scikit-learn.org/stable/modules/generated/sklearn.cluster.DBSCAN.html).  \n",
    "c) [AgglomerativeClustering](https://scikit-learn.org/stable/modules/generated/sklearn.cluster.AgglomerativeClustering.html). Рассмотрите число кластеров K = 2, 5, 10, 20."
   ]
  },
  {
   "cell_type": "markdown",
   "id": "bb97937b-13d7-44f9-b574-d2668355c8d9",
   "metadata": {},
   "source": [
    "#### 3. Для каждой модели оцените потери от уменьшения цветов при помощи метрики SSIM."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "id": "b8397107-dd33-4142-b9a3-fe38f1867179",
   "metadata": {},
   "outputs": [],
   "source": [
    "from sklearn.cluster import KMeans"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "id": "68154357-5690-417b-9abe-fe8bd8a96871",
   "metadata": {},
   "outputs": [],
   "source": [
    "def kmeans_clustering(x_train, i, r_stat):\n",
    "    kmeans = KMeans(n_clusters=i, random_state=r_stat, n_init=\"auto\").fit(x_train)\n",
    "    centers = kmeans.cluster_centers_ # founded centroids\n",
    "    n = centers.shape[0]\n",
    "    centers = np.c_[centers, np.arange(n)] # assigned count from 0 to n\n",
    "    centroids = pd.DataFrame(centers, columns =['B','G','R','N_Claster']).round().astype(np.uint8) # moved to dataframe format and one format for values\n",
    "    clasters = kmeans.predict(X_train).round().astype(np.uint8) # predicted clusters numbers\n",
    "    data = pd.concat([X_train,pd.DataFrame(clasters, columns=['N_Claster'])], axis=1) # moved to dataframe format\n",
    "    output = pd.merge(data,centroids, how='left', on = 'N_Claster') # merged with centroids' dataframe\n",
    "    b5 = output['B_y'].to_numpy().reshape((-1,128)) # chosen RGB dimension with further traansformation to required format\n",
    "    g5 = output['G_y'].to_numpy().reshape((-1,128))\n",
    "    r5 = output['R_y'].to_numpy().reshape((-1,128))\n",
    "    return cv2.merge([b5,g5,r5])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "id": "ae14a345-185f-48db-a23a-77cc016002fe",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 700x500 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 700x500 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 700x500 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 700x500 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "r_stat = 0\n",
    "for i in (2,5,10,20):\n",
    "    n_clusters = i\n",
    "    new_test = kmeans_clustering(X_train, n_clusters, r_stat)\n",
    "    acc = ssim(img, new_test, channel_axis=2)\n",
    "    draw_picture(new_test, f'Число кластеров = {n_clusters}, SSIM = {acc:.3f}')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "id": "dbfaa686-e089-42d4-8b63-8699da0fffdf",
   "metadata": {},
   "outputs": [],
   "source": [
    "from sklearn.cluster import DBSCAN"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "id": "d0a3ff17-bc4d-48f4-81f9-749e4a43ea21",
   "metadata": {},
   "outputs": [],
   "source": [
    "cl_DBSCAN = DBSCAN(eps=5).fit(X_train)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "id": "2ea2cc55-cd14-4ed7-a214-d251573b4e18",
   "metadata": {},
   "outputs": [],
   "source": [
    "labels = cl_DBSCAN.labels_"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "id": "90857b85-9d82-4e9a-ba75-5bf2c2f4b92f",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Estimated number of clusters: 49\n",
      "Estimated number of noise points: 1300\n"
     ]
    }
   ],
   "source": [
    "# Number of clusters in labels, ignoring noise if present.\n",
    "n_clusters_ = len(set(labels)) - (1 if -1 in labels else 0)\n",
    "n_noise_ = list(labels).count(-1)\n",
    "\n",
    "print(\"Estimated number of clusters: %d\" % n_clusters_)\n",
    "print(\"Estimated number of noise points: %d\" % n_noise_)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "id": "01e6b1de-b379-424d-a9e5-2bee01030eef",
   "metadata": {},
   "outputs": [],
   "source": [
    "data = pd.concat([X_train,pd.DataFrame(labels, columns=['N_Cluster'])], axis=1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "id": "8bc66db7-7f5f-4120-a475-49489e104191",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>B</th>\n",
       "      <th>G</th>\n",
       "      <th>R</th>\n",
       "      <th>N_Cluster</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>253</td>\n",
       "      <td>255</td>\n",
       "      <td>253</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>253</td>\n",
       "      <td>255</td>\n",
       "      <td>253</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>253</td>\n",
       "      <td>255</td>\n",
       "      <td>255</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>251</td>\n",
       "      <td>255</td>\n",
       "      <td>255</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>251</td>\n",
       "      <td>255</td>\n",
       "      <td>255</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "     B    G    R  N_Cluster\n",
       "0  253  255  253          0\n",
       "1  253  255  253          0\n",
       "2  253  255  255          0\n",
       "3  251  255  255          0\n",
       "4  251  255  255          0"
      ]
     },
     "execution_count": 19,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "data.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "id": "97529648-37ff-4bf5-8dbe-e16e4c04c246",
   "metadata": {},
   "outputs": [],
   "source": [
    "# поскольку центроидов в этом методе нет, то выберем центроидами средние значения\n",
    "clusters_mean = data.groupby('N_Cluster', as_index=False).\\\n",
    "                agg({'B':'mean','G':'mean','R':'mean'}).round().astype(np.uint8)\n",
    "clusters_mean.N_Cluster[0] = -1"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "id": "2f873dfc-7dec-4055-95f9-42ef64badb18",
   "metadata": {},
   "outputs": [],
   "source": [
    "output = pd.merge(data,clusters_mean, how='left', on = 'N_Cluster')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "id": "704027fb-48de-40d7-b5f8-2bc38f2fbd47",
   "metadata": {},
   "outputs": [],
   "source": [
    " # merged with centroids' dataframe\n",
    "bb = output['B_y'].to_numpy().reshape((-1,128))\n",
    "gg = output['G_y'].to_numpy().reshape((-1,128))\n",
    "rr = output['R_y'].to_numpy().reshape((-1,128))\n",
    "new_ = cv2.merge([bb,gg,rr])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "id": "493e7e75-b97f-48f3-8197-73b049f340ad",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 700x500 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "acc = ssim(img, new_, channel_axis=2)\n",
    "draw_picture(new_, f'DBSCAN, SSIM = {acc:.3f}')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "id": "668ba674-bc06-483f-ae28-5313387c2e3c",
   "metadata": {},
   "outputs": [],
   "source": [
    "from sklearn.cluster import AgglomerativeClustering as AC"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "id": "d6666701-f921-4dae-b12c-164a600d9429",
   "metadata": {},
   "outputs": [],
   "source": [
    "def ac_clustering(x_train, n):\n",
    "    cl_AC = AC(n_clusters = n).fit(x_train)\n",
    "    labels_ac = cl_AC.labels_\n",
    "    data_ac = pd.concat([X_train,pd.DataFrame(labels_ac, columns=['N_Cluster'])], axis=1)\n",
    "    # поскольку центроидов в этом методе нет, то выберем центроидами средние значения\n",
    "    clusters_mean_ac = data_ac.groupby('N_Cluster', as_index=False).\\\n",
    "                agg({'B':'mean','G':'mean','R':'mean'}).round().astype(np.uint8)\n",
    "    output_ac = pd.merge(data_ac,clusters_mean_ac, how='left', on = 'N_Cluster')\n",
    "     # merged with centroids' dataframe\n",
    "    ba = output_ac['B_y'].to_numpy().reshape((-1,128))\n",
    "    ga = output_ac['G_y'].to_numpy().reshape((-1,128))\n",
    "    ra = output_ac['R_y'].to_numpy().reshape((-1,128))\n",
    "    return cv2.merge([ba,ga,ra])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "id": "8e487211-5f09-4a8c-b1d2-9bbdc7d26939",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 700x500 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 700x500 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 700x500 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 700x500 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "for i in (2,5,10,20):\n",
    "    n = i\n",
    "    new_ac = ac_clustering(X_train, n)\n",
    "    acc = ssim(img, new_ac, channel_axis=2)\n",
    "    draw_picture(new_ac, f'AC method. Число кластеров = {n}, SSIM = {acc:.3f}')"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "9d79c890-1d9c-4839-b8f3-89bd3c32aab2",
   "metadata": {},
   "source": [
    "#### 4. Сформулируйте выводы по проделанной работе.  \n",
    "a) Какая модель показала лучший результат?  \n",
    "Для получения зачета по этому домашнему заданию минимально должно быть реализовано обучение трех моделей с одним и тем же значением кластеров, где такой параметр присутствует."
   ]
  },
  {
   "cell_type": "markdown",
   "id": "f91ecd7c-bfd9-498f-bac7-163d05c5896f",
   "metadata": {},
   "source": [
    "### **Ответ:** В этом конкретном случае лучший результат (SSIM = 0.841) был получен при помощи KMeans при выборе 20 кластеров."
   ]
  }
 ],
 "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": "3.11.4"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}