prototype_clustering.ipynb

{
  "cells": [
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "B1K8vvvSjuwo",
        "outputId": "dc867b8b-9bee-4059-c90e-6066679be86f"
      },
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Looking in indexes: https://pypi.org/simple, https://us-python.pkg.dev/colab-wheels/public/simple/\n",
            "Requirement already satisfied: scikit-learn-extra in /usr/local/lib/python3.7/dist-packages (0.2.0)\n",
            "Requirement already satisfied: scipy>=0.19.1 in /usr/local/lib/python3.7/dist-packages (from scikit-learn-extra) (1.7.3)\n",
            "Requirement already satisfied: numpy>=1.13.3 in /usr/local/lib/python3.7/dist-packages (from scikit-learn-extra) (1.21.6)\n",
            "Requirement already satisfied: scikit-learn>=0.23.0 in /usr/local/lib/python3.7/dist-packages (from scikit-learn-extra) (1.0.2)\n",
            "Requirement already satisfied: joblib>=0.11 in /usr/local/lib/python3.7/dist-packages (from scikit-learn>=0.23.0->scikit-learn-extra) (1.2.0)\n",
            "Requirement already satisfied: threadpoolctl>=2.0.0 in /usr/local/lib/python3.7/dist-packages (from scikit-learn>=0.23.0->scikit-learn-extra) (3.1.0)\n",
            "Looking in indexes: https://pypi.org/simple, https://us-python.pkg.dev/colab-wheels/public/simple/\n",
            "Requirement already satisfied: pyclustering in /usr/local/lib/python3.7/dist-packages (0.10.1.2)\n",
            "Requirement already satisfied: scipy>=1.1.0 in /usr/local/lib/python3.7/dist-packages (from pyclustering) (1.7.3)\n",
            "Requirement already satisfied: numpy>=1.15.2 in /usr/local/lib/python3.7/dist-packages (from pyclustering) (1.21.6)\n",
            "Requirement already satisfied: matplotlib>=3.0.0 in /usr/local/lib/python3.7/dist-packages (from pyclustering) (3.2.2)\n",
            "Requirement already satisfied: Pillow>=5.2.0 in /usr/local/lib/python3.7/dist-packages (from pyclustering) (7.1.2)\n",
            "Requirement already satisfied: cycler>=0.10 in /usr/local/lib/python3.7/dist-packages (from matplotlib>=3.0.0->pyclustering) (0.11.0)\n",
            "Requirement already satisfied: kiwisolver>=1.0.1 in /usr/local/lib/python3.7/dist-packages (from matplotlib>=3.0.0->pyclustering) (1.4.4)\n",
            "Requirement already satisfied: python-dateutil>=2.1 in /usr/local/lib/python3.7/dist-packages (from matplotlib>=3.0.0->pyclustering) (2.8.2)\n",
            "Requirement already satisfied: pyparsing!=2.0.4,!=2.1.2,!=2.1.6,>=2.0.1 in /usr/local/lib/python3.7/dist-packages (from matplotlib>=3.0.0->pyclustering) (3.0.9)\n",
            "Requirement already satisfied: typing-extensions in /usr/local/lib/python3.7/dist-packages (from kiwisolver>=1.0.1->matplotlib>=3.0.0->pyclustering) (4.1.1)\n",
            "Requirement already satisfied: six>=1.5 in /usr/local/lib/python3.7/dist-packages (from python-dateutil>=2.1->matplotlib>=3.0.0->pyclustering) (1.15.0)\n"
          ]
        }
      ],
      "source": [
        "# installing library for using KMedoids\n",
        "!pip install scikit-learn-extra\n",
        "# installing library for using clarans\n",
        "!pip3 install pyclustering"
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "# **About dataset**\n",
        "Measurements of geometrical properties of kernels belonging to three different varieties of wheat. A soft X-ray technique and GRAINS package were used to construct all seven, real-valued attributes.\n",
        "</br>\n",
        "Dataset link: https://archive.ics.uci.edu/ml/datasets/seeds"
      ],
      "metadata": {
        "id": "4zHW9of0ur6-"
      }
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "4tYawUOHm5DR"
      },
      "outputs": [],
      "source": [
        "#import all necessary libraries\n",
        "import pandas as pd\n",
        "import numpy as np\n",
        "import matplotlib.pyplot as plt\n",
        "from sklearn import metrics\n",
        "import warnings\n",
        "import seaborn as sns\n",
        "from sklearn_extra.cluster import KMedoids\n",
        "from sklearn.cluster import KMeans\n",
        "from sklearn.manifold import TSNE\n",
        "from sklearn import metrics\n",
        "from pyclustering.cluster.clarans import clarans\n",
        "warnings.filterwarnings('ignore')"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 206
        },
        "id": "fc889oe5m7Qd",
        "outputId": "5e4fc756-6c4b-4763-fe2f-bd6ac4bdc46d"
      },
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "       0      1       2      3      4      5      6  7\n",
              "0  15.26  14.84  0.8710  5.763  3.312  2.221  5.220  1\n",
              "1  14.88  14.57  0.8811  5.554  3.333  1.018  4.956  1\n",
              "2  14.29  14.09  0.9050  5.291  3.337  2.699  4.825  1\n",
              "3  13.84  13.94  0.8955  5.324  3.379  2.259  4.805  1\n",
              "4  16.14  14.99  0.9034  5.658  3.562  1.355  5.175  1"
            ],
            "text/html": [
              "\n",
              "  <div id=\"df-0090108a-374f-43d1-9b92-d7da0aafdab6\">\n",
              "    <div class=\"colab-df-container\">\n",
              "      <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>0</th>\n",
              "      <th>1</th>\n",
              "      <th>2</th>\n",
              "      <th>3</th>\n",
              "      <th>4</th>\n",
              "      <th>5</th>\n",
              "      <th>6</th>\n",
              "      <th>7</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>15.26</td>\n",
              "      <td>14.84</td>\n",
              "      <td>0.8710</td>\n",
              "      <td>5.763</td>\n",
              "      <td>3.312</td>\n",
              "      <td>2.221</td>\n",
              "      <td>5.220</td>\n",
              "      <td>1</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1</th>\n",
              "      <td>14.88</td>\n",
              "      <td>14.57</td>\n",
              "      <td>0.8811</td>\n",
              "      <td>5.554</td>\n",
              "      <td>3.333</td>\n",
              "      <td>1.018</td>\n",
              "      <td>4.956</td>\n",
              "      <td>1</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2</th>\n",
              "      <td>14.29</td>\n",
              "      <td>14.09</td>\n",
              "      <td>0.9050</td>\n",
              "      <td>5.291</td>\n",
              "      <td>3.337</td>\n",
              "      <td>2.699</td>\n",
              "      <td>4.825</td>\n",
              "      <td>1</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3</th>\n",
              "      <td>13.84</td>\n",
              "      <td>13.94</td>\n",
              "      <td>0.8955</td>\n",
              "      <td>5.324</td>\n",
              "      <td>3.379</td>\n",
              "      <td>2.259</td>\n",
              "      <td>4.805</td>\n",
              "      <td>1</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4</th>\n",
              "      <td>16.14</td>\n",
              "      <td>14.99</td>\n",
              "      <td>0.9034</td>\n",
              "      <td>5.658</td>\n",
              "      <td>3.562</td>\n",
              "      <td>1.355</td>\n",
              "      <td>5.175</td>\n",
              "      <td>1</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>\n",
              "      <button class=\"colab-df-convert\" onclick=\"convertToInteractive('df-0090108a-374f-43d1-9b92-d7da0aafdab6')\"\n",
              "              title=\"Convert this dataframe to an interactive table.\"\n",
              "              style=\"display:none;\">\n",
              "        \n",
              "  <svg xmlns=\"http://www.w3.org/2000/svg\" height=\"24px\"viewBox=\"0 0 24 24\"\n",
              "       width=\"24px\">\n",
              "    <path d=\"M0 0h24v24H0V0z\" fill=\"none\"/>\n",
              "    <path d=\"M18.56 5.44l.94 2.06.94-2.06 2.06-.94-2.06-.94-.94-2.06-.94 2.06-2.06.94zm-11 1L8.5 8.5l.94-2.06 2.06-.94-2.06-.94L8.5 2.5l-.94 2.06-2.06.94zm10 10l.94 2.06.94-2.06 2.06-.94-2.06-.94-.94-2.06-.94 2.06-2.06.94z\"/><path d=\"M17.41 7.96l-1.37-1.37c-.4-.4-.92-.59-1.43-.59-.52 0-1.04.2-1.43.59L10.3 9.45l-7.72 7.72c-.78.78-.78 2.05 0 2.83L4 21.41c.39.39.9.59 1.41.59.51 0 1.02-.2 1.41-.59l7.78-7.78 2.81-2.81c.8-.78.8-2.07 0-2.86zM5.41 20L4 18.59l7.72-7.72 1.47 1.35L5.41 20z\"/>\n",
              "  </svg>\n",
              "      </button>\n",
              "      \n",
              "  <style>\n",
              "    .colab-df-container {\n",
              "      display:flex;\n",
              "      flex-wrap:wrap;\n",
              "      gap: 12px;\n",
              "    }\n",
              "\n",
              "    .colab-df-convert {\n",
              "      background-color: #E8F0FE;\n",
              "      border: none;\n",
              "      border-radius: 50%;\n",
              "      cursor: pointer;\n",
              "      display: none;\n",
              "      fill: #1967D2;\n",
              "      height: 32px;\n",
              "      padding: 0 0 0 0;\n",
              "      width: 32px;\n",
              "    }\n",
              "\n",
              "    .colab-df-convert:hover {\n",
              "      background-color: #E2EBFA;\n",
              "      box-shadow: 0px 1px 2px rgba(60, 64, 67, 0.3), 0px 1px 3px 1px rgba(60, 64, 67, 0.15);\n",
              "      fill: #174EA6;\n",
              "    }\n",
              "\n",
              "    [theme=dark] .colab-df-convert {\n",
              "      background-color: #3B4455;\n",
              "      fill: #D2E3FC;\n",
              "    }\n",
              "\n",
              "    [theme=dark] .colab-df-convert:hover {\n",
              "      background-color: #434B5C;\n",
              "      box-shadow: 0px 1px 3px 1px rgba(0, 0, 0, 0.15);\n",
              "      filter: drop-shadow(0px 1px 2px rgba(0, 0, 0, 0.3));\n",
              "      fill: #FFFFFF;\n",
              "    }\n",
              "  </style>\n",
              "\n",
              "      <script>\n",
              "        const buttonEl =\n",
              "          document.querySelector('#df-0090108a-374f-43d1-9b92-d7da0aafdab6 button.colab-df-convert');\n",
              "        buttonEl.style.display =\n",
              "          google.colab.kernel.accessAllowed ? 'block' : 'none';\n",
              "\n",
              "        async function convertToInteractive(key) {\n",
              "          const element = document.querySelector('#df-0090108a-374f-43d1-9b92-d7da0aafdab6');\n",
              "          const dataTable =\n",
              "            await google.colab.kernel.invokeFunction('convertToInteractive',\n",
              "                                                     [key], {});\n",
              "          if (!dataTable) return;\n",
              "\n",
              "          const docLinkHtml = 'Like what you see? Visit the ' +\n",
              "            '<a target=\"_blank\" href=https://colab.research.google.com/notebooks/data_table.ipynb>data table notebook</a>'\n",
              "            + ' to learn more about interactive tables.';\n",
              "          element.innerHTML = '';\n",
              "          dataTable['output_type'] = 'display_data';\n",
              "          await google.colab.output.renderOutput(dataTable, element);\n",
              "          const docLink = document.createElement('div');\n",
              "          docLink.innerHTML = docLinkHtml;\n",
              "          element.appendChild(docLink);\n",
              "        }\n",
              "      </script>\n",
              "    </div>\n",
              "  </div>\n",
              "  "
            ]
          },
          "metadata": {},
          "execution_count": 21
        }
      ],
      "source": [
        "# load the the seed dataset \n",
        "seed=pd.read_csv('seeds.csv',header=None)\n",
        "#Print the Dataframe\n",
        "seed.head()"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "6tBImEgrLNLq",
        "outputId": "f900fa04-d05e-4f19-c60d-a2f9bf973b31"
      },
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "array([1, 2, 3])"
            ]
          },
          "metadata": {},
          "execution_count": 22
        }
      ],
      "source": [
        "# Unique class labels\n",
        "seed[7].unique()"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "I22eLQ3XLVLD"
      },
      "outputs": [],
      "source": [
        "#mapping the seed labels\n",
        "mp={1:0,2:1,3:2}\n",
        "seed[7]=seed[7].map(mp)"
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "print(f'Duplicate rows in the datsaset :{seed.duplicated().sum()}\\n') "
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "MdzBFDintomn",
        "outputId": "d61fd04c-5c03-4963-eb38-91fa4647fc56"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Duplicate rows in the datsaset :0\n",
            "\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "h3zndWUcLXl3",
        "outputId": "a488dadf-5b07-4fb5-a568-9a4c26900f7e"
      },
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "(210, 8)"
            ]
          },
          "metadata": {},
          "execution_count": 25
        }
      ],
      "source": [
        "# Displaying the number of instances and attributes\n",
        "seed.shape"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "R2-TgzHILZJB",
        "outputId": "e4271791-1263-4743-bb3a-9400f043a76e"
      },
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "0    0\n",
              "1    0\n",
              "2    0\n",
              "3    0\n",
              "4    0\n",
              "5    0\n",
              "6    0\n",
              "7    0\n",
              "dtype: int64"
            ]
          },
          "metadata": {},
          "execution_count": 26
        }
      ],
      "source": [
        "# Total null values in each column\n",
        "seed.isna().sum()"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "2EracG0DLd9o"
      },
      "outputs": [],
      "source": [
        "#extracting feature columns from the tumor dataframe\n",
        "feature_cols_seed=seed.columns.tolist()[:-1]"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "dvU-21RkeOgA"
      },
      "source": [
        "# **Limitations of KMeans** \n",
        "(i) **Clustering Outliers**\n",
        "\n",
        "* The KMeans algorithm relies on concept of calculating the centroid during every iteration by finding out the mean of all the points within the cluster at that time. Also in KMeans algorithm, each point must belong to one cluster.\n",
        "* Hence the presence of outliers can substantially change the mean and adversely affect the clustering process. This is a limitation of prototype clustering.\n",
        "\n",
        "\n",
        "(ii) **Scaling with number of dimensions**\n",
        "* The KMeans algorithm performs very badly for datasets with a large no of dimensions. The reason for this is that the sparsity of the dataspace increases exponentially with increasing number of dimensions.\n",
        "\n",
        "* Also with increasing number of dimensions,the notion of proximity turns out to be very ineffective due to highly sparse high dimensional data spaces. In such a data space, all points appear to be equidistant. Hence prototype clustering doesn't work well with high dimensional data.\n"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "O8y6gfRpSboY"
      },
      "source": [
        "# **Algorithms addressing limitations of KMeans**\n",
        "\n",
        "(i) **Clustering Outliers** \n",
        "* The prototype clustering method that tries to overcome the limitation of KMeans for data with outliers is called **KMedoids**. In this variant of KMeans, the cluster centres are not random points in the data space rather they are datapoints themselves. \n",
        "* Clustering is done by identifying the most centrally located data point as the cluster centre also called medoids.\n",
        "* Noise points can never be centrally located data points and hence can never become the cluster centres or medoids. Hence noise becomes secluded from the clusters formed and so KMedoids is robust to outliers. \n",
        "\n",
        "(ii) **Scaling with number of dimensions** \n",
        "\n",
        "* Clustering Large Applications based upon Randomized Search (CLARANS). The KMedoid clustering technique removes the limitation of KMeans of getting affected by oultier in the input data. But KMedoid is computationally very costly for large values of K and high dimensional data.\n",
        "* The CLARA algorithm is an extension of KMedoids and instead of using the entire input data it uses only random sample of it. But it does not peroform well if one or more medoid in the sampled input data is far away from the actual medoid in the input data. \n",
        "* CLARANS takes care of the limitation of KMedoid and CLARA by maintaining  a balance between computational cost and the influence of data sampling on clustering. CLARA does the sampling at beginning of the process however CLARANS does sampling of neighbors in each iteration.  \n",
        "\n",
        "\n"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "ChhRto7SZ4mL"
      },
      "source": [
        "# **Clustering using KMedoid, CLARANS and comparison with KMeans**"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "7imdJuGXZpRi"
      },
      "source": [
        "## **Visualization of given class labels**"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 552
        },
        "id": "_ldqDRCoL-Cv",
        "outputId": "b59bb9ec-7046-4801-c149-37a4086d9f1d"
      },
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 576x648 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {
            "needs_background": "light"
          }
        }
      ],
      "source": [
        "\n",
        "\n",
        "\n",
        "#Performing tsne on the dataset which will help in cluster visualization\n",
        "tsne = TSNE(n_components = 2, random_state = 0)\n",
        "tsne_euc = tsne.fit_transform(seed[feature_cols_seed])\n",
        "seed_labels= seed[7].values\n",
        "seed_labels=np.array(seed_labels).reshape(-1,1)\n",
        "\n",
        "\n",
        "# creating a new tsne dataframe which help us in plotting the groundtruth and resulting cluter later on \n",
        "tsne_data = np.concatenate((tsne_euc, seed_labels), axis=1)\n",
        "tsne_df = pd.DataFrame(data = tsne_data,\n",
        "     columns =(\"Dim_1\", \"Dim_2\", \"Seed Label\"))\n",
        "\n",
        "\n",
        "colors = sns.color_palette()\n",
        "modes_gt=tsne_df['Seed Label'].unique()\n",
        "palette_gt = {mode: color for mode, color in zip(modes_gt, colors)}\n",
        "\n",
        "\n",
        "plt.figure(figsize = (8, 9))\n",
        "sns.scatterplot(x=\"Dim_1\", y=\"Dim_2\", data=tsne_df, hue='Seed Label',palette=palette_gt,alpha=0.8);"
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "## **Clustering and Visualization of dataset**"
      ],
      "metadata": {
        "id": "bpYlFpjLzp8U"
      }
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 1000
        },
        "id": "El9U3VQUUtl3",
        "outputId": "b0a36831-4d28-4e6c-df98-edb2aaecc3db"
      },
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 1152x864 with 3 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {
            "needs_background": "light"
          }
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "****************************\n",
            "KMedoids\n",
            "Intrinsic Measure:\n",
            "\n",
            "Silhouette Coefficient: 0.480\n",
            "Calinski and Harabasz score: 243.876\n",
            "Davies-Bouldin score: 0.894\n",
            "\n",
            "Extrinsic Measure:\n",
            "\n",
            "Random index: 0.876\n",
            "Fowlkes-Mallows score: 0.815\n",
            "Adjusted Random index: 0.722\n",
            "Adjused mutual Info score: 0.698\n",
            "****************************\n",
            "CLARANS\n",
            "Intrinsic Measure:\n",
            "\n",
            "Silhouette Coefficient: 0.480\n",
            "Calinski and Harabasz score: 244.790\n",
            "Davies-Bouldin score: 0.855\n",
            "\n",
            "Extrinsic Measure:\n",
            "\n",
            "Random index: 0.832\n",
            "Fowlkes-Mallows score: 0.755\n",
            "Adjusted Random index: 0.628\n",
            "Adjused mutual Info score: 0.627\n",
            "****************************\n",
            "KMeans\n",
            "Intrinsic Measure:\n",
            "\n",
            "Silhouette Coefficient: 0.493\n",
            "Calinski and Harabasz score: 258.538\n",
            "Davies-Bouldin score: 0.855\n",
            "\n",
            "Extrinsic Measure:\n",
            "\n",
            "Random index: 0.862\n",
            "Fowlkes-Mallows score: 0.793\n",
            "Adjusted Random index: 0.689\n",
            "Adjused mutual Info score: 0.657\n"
          ]
        }
      ],
      "source": [
        "from sklearn.preprocessing import normalize, StandardScaler\n",
        "\n",
        "\n",
        "# z-score normalization on data\n",
        "seed_std= StandardScaler().fit_transform(seed[feature_cols_seed])\n",
        "# converting each feature vector into unit vector\n",
        "normalized_df_seed=normalize(seed_std)\n",
        "\n",
        "\n",
        "tsne_df_clus=tsne_df.copy()\n",
        "\n",
        "# KMedoids Clustering of seeds dataset (dataset 3)\n",
        "clusterer_kmed=KMedoids(n_clusters=3, random_state=0,method='pam',metric='manhattan',init='k-medoids++').fit(normalized_df_seed)\n",
        "tsne_df_clus['KMedoids cluster labels']=clusterer_kmed.labels_\n",
        "\n",
        "\n",
        "\n",
        "\n",
        "# Clustering using ClARANS of seeds dataset (dataset 3)\n",
        "d={}\n",
        "# The choice of the parameters is motivated by the fact\n",
        "#that high computing resources are required for CALARNS\n",
        "#hence lower values of parameters have been chosen in the absence of high computing machine\n",
        "clarans_instance = clarans(normalized_df_seed, number_clusters=3, numlocal=4, maxneighbor=3)\n",
        "clarans_instance.process()\n",
        "d['clusters'] = clarans_instance.get_clusters()\n",
        "d['medoids'] = clarans_instance.get_medoids()\n",
        "\n",
        "# clarans return list of index of data points in each cluster \n",
        "# code to map clarans clsuter label to the existing dataframe\n",
        "indices_0=d['clusters'][0]\n",
        "indices_1=d['clusters'][1]\n",
        "indices_2=d['clusters'][2]\n",
        "tsne_df_clus['CLARANS cluster labels']=[-1]* len(tsne_df_clus)\n",
        "\n",
        "tsne_df_clus.iloc[tsne_df_clus.index.isin(indices_0),[4]]=0\n",
        "tsne_df_clus.iloc[tsne_df_clus.index.isin(indices_1),[4]]=1\n",
        "tsne_df_clus.iloc[tsne_df_clus.index.isin(indices_2),[4]]=2\n",
        "\n",
        "\n",
        "\n",
        "#Clustering using KMeans of seeds dataset (dataset 3)\n",
        "clusterer_kmeans=KMeans(n_clusters=3).fit(normalized_df_seed)\n",
        "tsne_df_clus['KMeans cluster labels']=clusterer_kmeans.labels_\n",
        "\n",
        "\n",
        "\n",
        "\n",
        "# Visualization of cluster labels obtained using KMedoids, CLARANS and  KMeans\n",
        "colors = sns.color_palette()\n",
        "\n",
        "modes_kmed=tsne_df_clus['KMedoids cluster labels'].unique()\n",
        "palette_kmed = {mode: color for mode, color in zip(modes_kmed, colors)}\n",
        "\n",
        "modes_clus_clara=tsne_df_clus['CLARANS cluster labels'].unique()\n",
        "palette_clus_clara = {mode: color for mode, color in zip(modes_clus_clara, colors)}\n",
        "\n",
        "modes_clus_kmean=tsne_df_clus['KMeans cluster labels'].unique()\n",
        "palette_clus_kmean = {mode: color for mode, color in zip(modes_clus_kmean, colors)}\n",
        "\n",
        "plt.figure(figsize = (16, 12))\n",
        "ax = plt.subplot(1,3,1)\n",
        "sns.scatterplot(x=\"Dim_1\", y=\"Dim_2\", data=tsne_df_clus, ax= ax,hue='KMedoids cluster labels',palette=palette_kmed,alpha=0.8);\n",
        "ax = plt.subplot(1,3,2)\n",
        "sns.scatterplot(x=\"Dim_1\", y=\"Dim_2\", data=tsne_df_clus, ax= ax, hue='CLARANS cluster labels',palette=palette_clus_clara,alpha=0.8);\n",
        "ax = plt.subplot(1,3,3)\n",
        "sns.scatterplot(x=\"Dim_1\", y=\"Dim_2\", data=tsne_df_clus, ax= ax, hue='KMeans cluster labels',palette=palette_clus_kmean,alpha=0.8);\n",
        "plt.show()\n",
        "\n",
        "\n",
        "\n",
        "\n",
        "\n",
        "# computing various cluster performance measures\n",
        "\n",
        "\n",
        "\n",
        "# For KMedoids\n",
        "\n",
        "y_true=tsne_df_clus['Seed Label'].values.tolist()\n",
        "y_pred=tsne_df_clus['KMedoids cluster labels'].values.tolist()\n",
        "\n",
        "\n",
        "# calculating intrinsic measures \n",
        "SIL_ward=metrics.silhouette_score(normalized_df_seed, y_pred)\n",
        "CH_ward=metrics.calinski_harabasz_score(normalized_df_seed, y_pred)\n",
        "DB_ward=metrics.davies_bouldin_score( normalized_df_seed,  y_pred)\n",
        "\n",
        "\n",
        "# calculating extrinsic measures\n",
        "RI_ward=metrics.rand_score(y_true,y_pred)\n",
        "FM_ward=metrics.fowlkes_mallows_score(y_true,y_pred)\n",
        "ARS_ward=metrics.adjusted_rand_score(y_true,y_pred)\n",
        "AMS_ward=metrics.adjusted_mutual_info_score(y_true,y_pred)\n",
        "\n",
        "print('****************************')\n",
        "print(f'KMedoids')\n",
        "print('Intrinsic Measure:\\n')\n",
        "print(f'Silhouette Coefficient: {SIL_ward:.3f}')\n",
        "print(f'Calinski and Harabasz score: {CH_ward:.3f}')\n",
        "print(f'Davies-Bouldin score: {DB_ward:.3f}')\n",
        "print('')\n",
        " \n",
        "print('Extrinsic Measure:\\n')\n",
        "\n",
        "print(f'Random index: {RI_ward:.3f}')\n",
        "print(f'Fowlkes-Mallows score: {FM_ward:.3f}')\n",
        "print(f'Adjusted Random index: {ARS_ward:.3f}')\n",
        "print(f'Adjused mutual Info score: {AMS_ward:.3f}')\n",
        "\n",
        "\n",
        "# For CLARANS\n",
        "\n",
        "y_true=tsne_df_clus['Seed Label'].values.tolist()\n",
        "y_pred=tsne_df_clus['CLARANS cluster labels'].values.tolist()\n",
        "\n",
        "\n",
        "# calculating intrinsic measures \n",
        "SIL_ward=metrics.silhouette_score(normalized_df_seed, y_pred)\n",
        "CH_ward=metrics.calinski_harabasz_score(normalized_df_seed, y_pred)\n",
        "DB_ward=metrics.davies_bouldin_score( normalized_df_seed,  y_pred)\n",
        "\n",
        "\n",
        "# calculating extrinsic measures\n",
        "RI_ward=metrics.rand_score(y_true,y_pred)\n",
        "FM_ward=metrics.fowlkes_mallows_score(y_true,y_pred)\n",
        "ARS_ward=metrics.adjusted_rand_score(y_true,y_pred)\n",
        "AMS_ward=metrics.adjusted_mutual_info_score(y_true,y_pred)\n",
        "\n",
        "\n",
        "print('****************************')\n",
        "print(f'CLARANS')\n",
        "print('Intrinsic Measure:\\n')\n",
        "print(f'Silhouette Coefficient: {SIL_ward:.3f}')\n",
        "print(f'Calinski and Harabasz score: {CH_ward:.3f}')\n",
        "print(f'Davies-Bouldin score: {DB_ward:.3f}')\n",
        "print('')\n",
        " \n",
        "print('Extrinsic Measure:\\n')\n",
        "\n",
        "print(f'Random index: {RI_ward:.3f}')\n",
        "print(f'Fowlkes-Mallows score: {FM_ward:.3f}')\n",
        "print(f'Adjusted Random index: {ARS_ward:.3f}')\n",
        "print(f'Adjused mutual Info score: {AMS_ward:.3f}')\n",
        "\n",
        "\n",
        "\n",
        "# For KMeans\n",
        "\n",
        "y_true=tsne_df_clus['Seed Label'].values.tolist()\n",
        "y_pred=tsne_df_clus['KMeans cluster labels'].values.tolist()\n",
        "\n",
        "\n",
        "# calculating intrinsic measures \n",
        "SIL_ward=metrics.silhouette_score(normalized_df_seed, y_pred)\n",
        "CH_ward=metrics.calinski_harabasz_score(normalized_df_seed, y_pred)\n",
        "DB_ward=metrics.davies_bouldin_score( normalized_df_seed,  y_pred)\n",
        "\n",
        "\n",
        "# calculating extrinsic measures\n",
        "RI_ward=metrics.rand_score(y_true,y_pred)\n",
        "FM_ward=metrics.fowlkes_mallows_score(y_true,y_pred)\n",
        "ARS_ward=metrics.adjusted_rand_score(y_true,y_pred)\n",
        "AMS_ward=metrics.adjusted_mutual_info_score(y_true,y_pred)\n",
        "\n",
        "\n",
        "print('****************************')\n",
        "print(f'KMeans')\n",
        "print('Intrinsic Measure:\\n')\n",
        "print(f'Silhouette Coefficient: {SIL_ward:.3f}')\n",
        "print(f'Calinski and Harabasz score: {CH_ward:.3f}')\n",
        "print(f'Davies-Bouldin score: {DB_ward:.3f}')\n",
        "print('')\n",
        " \n",
        "print('Extrinsic Measure:\\n')\n",
        "\n",
        "print(f'Random index: {RI_ward:.3f}')\n",
        "print(f'Fowlkes-Mallows score: {FM_ward:.3f}')\n",
        "print(f'Adjusted Random index: {ARS_ward:.3f}')\n",
        "print(f'Adjused mutual Info score: {AMS_ward:.3f}')\n",
        "\n",
        "\n",
        "\n",
        "\n",
        "\n"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "ac5iE1c8i3kO"
      },
      "source": [
        "## Standard behaviour of the cluster validity indices used\n",
        "* The silhouette ranges from −1 to +1, where a high value indicates that the object is well matched to its own cluster and poorly matched to neighboring clusters.\n",
        "* The minimum value for Davies-Bouldin score is 0, with lower values indicating better clustering.\n",
        "* The higher the Calinski and Harabasz score , the better the performances. \n",
        "* The Rand Index may only yield a value between 0 and 1 with higher value indicating better clustering. \n",
        "\n",
        "* Fowlkes-Mallows score ranges from 0 to 1. A high value indicates a good similarity between two clusters."
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "99eu2HuD8rst"
      },
      "source": [
        "## **Comparison of the results of KMedoids, CLARANS and KMeans clustering on the  based on cluster validity indices**\n",
        "* Extrinsic measures perform better for KMedoid, CLARANS and KMeans report the same values for extrinsic measures.\n",
        "* For intrinsic measures KMeans perform slightly better than KMedoid and CLARA.\n",
        "* CLARANS perform better than KMedoids in intrinsic measures while KMedoid peroforms better in extrinsic measures.\n",
        "* It can be concluded that all the three algorithms perform well on dataset 3.\n",
        "\n",
        "\n"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "Ci-iB1t05N2O"
      },
      "source": [
        "# **Strategy adopted for Hyperparameter tuning**\n"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "0gsbLAEQjBTH"
      },
      "source": [
        "For each dataset the hyperparameter tuning in case of KMedoid is done by varying the following input hyperparameters:\n",
        "\n",
        "* **metric** :  The distance metric to use for calculating proximity between clusters. We are tuning for:\n",
        "   * euclidean\n",
        "   * cosine \n",
        "   * manhattan</br>\n",
        "* **init** : The medoid initialization method to be used. It has the following values:\n",
        "  * 'random’ selects n_clusters elements from the dataset. \n",
        "  * ‘heuristic’ picks the n_clusters points with the smallest sum distance to every other point. \n",
        "  * ‘k-medoids++’ follows an approach based on k-means++ , and in general, gives initial medoids which are more separated than those generated by the other methods. \n",
        "  * ‘build’ is a greedy initialization of the medoids used in the original PAM algorithm "
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "a86VaM-dkRIH"
      },
      "source": [
        "The hyperparameters availabe in CLARANS are:\n",
        "* **number_clusters** : the number of clusters to create in the data\n",
        "* **maxneighbor**: maximum number of neighbors examined\n",
        "* **numlocal** : the number of local minima obtained (numLocal)\n",
        "\n",
        "\n",
        "\n",
        "\n",
        "\n",
        "\n",
        "\n",
        "   "
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "zg1MD6wh5ezW"
      },
      "source": [
        "## **Hyperparameter tuning for Kmedoids**"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "ma2Eo8T5jAb9",
        "outputId": "3ee776ab-3659-401b-b34a-29692c09aa33"
      },
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "****************************\n",
            "euclidean:random\n",
            "Intrinsic Measure:\n",
            "\n",
            "Silhouette Coefficient: 0.493\n",
            "Calinski and Harabasz score: 258.538\n",
            "Davies-Bouldin score: 0.855\n",
            "\n",
            "Extrinsic Measure:\n",
            "\n",
            "Random index: 0.862\n",
            "Fowlkes-Mallows score: 0.793\n",
            "\n",
            "****************************\n",
            "euclidean:heuristic\n",
            "Intrinsic Measure:\n",
            "\n",
            "Silhouette Coefficient: 0.493\n",
            "Calinski and Harabasz score: 258.538\n",
            "Davies-Bouldin score: 0.855\n",
            "\n",
            "Extrinsic Measure:\n",
            "\n",
            "Random index: 0.862\n",
            "Fowlkes-Mallows score: 0.793\n",
            "\n",
            "****************************\n",
            "euclidean:k-medoids++\n",
            "Intrinsic Measure:\n",
            "\n",
            "Silhouette Coefficient: 0.493\n",
            "Calinski and Harabasz score: 258.538\n",
            "Davies-Bouldin score: 0.855\n",
            "\n",
            "Extrinsic Measure:\n",
            "\n",
            "Random index: 0.862\n",
            "Fowlkes-Mallows score: 0.793\n",
            "\n",
            "****************************\n",
            "euclidean:build\n",
            "Intrinsic Measure:\n",
            "\n",
            "Silhouette Coefficient: 0.493\n",
            "Calinski and Harabasz score: 258.538\n",
            "Davies-Bouldin score: 0.855\n",
            "\n",
            "Extrinsic Measure:\n",
            "\n",
            "Random index: 0.862\n",
            "Fowlkes-Mallows score: 0.793\n",
            "\n",
            "****************************\n",
            "cosine:random\n",
            "Intrinsic Measure:\n",
            "\n",
            "Silhouette Coefficient: 0.491\n",
            "Calinski and Harabasz score: 255.798\n",
            "Davies-Bouldin score: 0.864\n",
            "\n",
            "Extrinsic Measure:\n",
            "\n",
            "Random index: 0.867\n",
            "Fowlkes-Mallows score: 0.801\n",
            "\n",
            "****************************\n",
            "cosine:heuristic\n",
            "Intrinsic Measure:\n",
            "\n",
            "Silhouette Coefficient: 0.491\n",
            "Calinski and Harabasz score: 255.798\n",
            "Davies-Bouldin score: 0.864\n",
            "\n",
            "Extrinsic Measure:\n",
            "\n",
            "Random index: 0.867\n",
            "Fowlkes-Mallows score: 0.801\n",
            "\n",
            "****************************\n",
            "cosine:k-medoids++\n",
            "Intrinsic Measure:\n",
            "\n",
            "Silhouette Coefficient: 0.491\n",
            "Calinski and Harabasz score: 255.798\n",
            "Davies-Bouldin score: 0.864\n",
            "\n",
            "Extrinsic Measure:\n",
            "\n",
            "Random index: 0.867\n",
            "Fowlkes-Mallows score: 0.801\n",
            "\n",
            "****************************\n",
            "cosine:build\n",
            "Intrinsic Measure:\n",
            "\n",
            "Silhouette Coefficient: 0.491\n",
            "Calinski and Harabasz score: 255.798\n",
            "Davies-Bouldin score: 0.864\n",
            "\n",
            "Extrinsic Measure:\n",
            "\n",
            "Random index: 0.867\n",
            "Fowlkes-Mallows score: 0.801\n",
            "\n",
            "****************************\n",
            "manhattan:random\n",
            "Intrinsic Measure:\n",
            "\n",
            "Silhouette Coefficient: 0.480\n",
            "Calinski and Harabasz score: 243.876\n",
            "Davies-Bouldin score: 0.894\n",
            "\n",
            "Extrinsic Measure:\n",
            "\n",
            "Random index: 0.876\n",
            "Fowlkes-Mallows score: 0.815\n",
            "\n",
            "****************************\n",
            "manhattan:heuristic\n",
            "Intrinsic Measure:\n",
            "\n",
            "Silhouette Coefficient: 0.480\n",
            "Calinski and Harabasz score: 243.876\n",
            "Davies-Bouldin score: 0.894\n",
            "\n",
            "Extrinsic Measure:\n",
            "\n",
            "Random index: 0.876\n",
            "Fowlkes-Mallows score: 0.815\n",
            "\n",
            "****************************\n",
            "manhattan:k-medoids++\n",
            "Intrinsic Measure:\n",
            "\n",
            "Silhouette Coefficient: 0.480\n",
            "Calinski and Harabasz score: 243.876\n",
            "Davies-Bouldin score: 0.894\n",
            "\n",
            "Extrinsic Measure:\n",
            "\n",
            "Random index: 0.876\n",
            "Fowlkes-Mallows score: 0.815\n",
            "\n",
            "****************************\n",
            "manhattan:build\n",
            "Intrinsic Measure:\n",
            "\n",
            "Silhouette Coefficient: 0.480\n",
            "Calinski and Harabasz score: 243.876\n",
            "Davies-Bouldin score: 0.894\n",
            "\n",
            "Extrinsic Measure:\n",
            "\n",
            "Random index: 0.876\n",
            "Fowlkes-Mallows score: 0.815\n",
            "\n"
          ]
        }
      ],
      "source": [
        "#using metric and init hyperparameters to tune for KMedoids\n",
        "met=['euclidean','cosine','manhattan'] # various values for metrics \n",
        "init=['random', 'heuristic', 'k-medoids++', 'build'] # various values for initialisation\n",
        "\n",
        "#creating a dictionasry to store metrics for various hyperparameters combination which will be later transofrmed into a dataframe\n",
        "dict_metrics_seed={}\n",
        "dict_metrics_seed['Metrics']=['Silhouette Coeff','Calinski and Harabasz score','Davies-Bouldin score','Random index','Fowlkes-Mallows score']\n",
        "\n",
        "for metric in met:\n",
        "  for initial in init:\n",
        "\n",
        "    clusterer = KMedoids(n_clusters=3, random_state=0,metric=metric,method='pam',init=initial).fit(normalized_df_seed)\n",
        "    y_true=seed_labels.tolist()\n",
        "    y_true=np.array(y_true).reshape(-1)\n",
        "    y_pred=np.array(clusterer.labels_).reshape(-1)\n",
        "    SIL_ward=metrics.silhouette_score(normalized_df_seed, clusterer.labels_)\n",
        "    CH_ward=metrics.calinski_harabasz_score(normalized_df_seed, clusterer.labels_)\n",
        "    DB_ward=metrics.davies_bouldin_score( normalized_df_seed,  clusterer.labels_)\n",
        "    \n",
        "    RI_ward=metrics.rand_score(y_true,y_pred)\n",
        "    FM_ward=metrics.fowlkes_mallows_score(y_true,y_pred)\n",
        "\n",
        "    dict_metrics_seed[metric+str(': ')+initial]=[SIL_ward,CH_ward,DB_ward,RI_ward,FM_ward]\n",
        "\n",
        "    print('****************************')\n",
        "    print(f'{metric}:{initial}')\n",
        "    print('Intrinsic Measure:\\n')\n",
        "    print(f'Silhouette Coefficient: {SIL_ward:.3f}')\n",
        "    print(f'Calinski and Harabasz score: {CH_ward:.3f}')\n",
        "    print(f'Davies-Bouldin score: {DB_ward:.3f}')\n",
        "    print('')\n",
        "    print('Extrinsic Measure:\\n')\n",
        "    print(f'Random index: {RI_ward:.3f}')\n",
        "    print(f'Fowlkes-Mallows score: {FM_ward:.3f}')\n",
        "\n",
        "    \n",
        "    print()"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 438
        },
        "id": "yrMQI_jkyc-9",
        "outputId": "4c65895e-208d-404d-8f6e-0305113914c1"
      },
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Presenting the results in a tabular form\n"
          ]
        },
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "                             euclidean: random  euclidean: heuristic  \\\n",
              "Metrics                                                                \n",
              "Silhouette Coeff                      0.493491              0.493491   \n",
              "Calinski and Harabasz score         258.538210            258.538210   \n",
              "Davies-Bouldin score                  0.854915              0.854915   \n",
              "Random index                          0.861609              0.861609   \n",
              "Fowlkes-Mallows score                 0.792906              0.792906   \n",
              "\n",
              "                             euclidean: k-medoids++  euclidean: build  \\\n",
              "Metrics                                                                 \n",
              "Silhouette Coeff                           0.493491          0.493491   \n",
              "Calinski and Harabasz score              258.538210        258.538210   \n",
              "Davies-Bouldin score                       0.854915          0.854915   \n",
              "Random index                               0.861609          0.861609   \n",
              "Fowlkes-Mallows score                      0.792906          0.792906   \n",
              "\n",
              "                             cosine: random  cosine: heuristic  \\\n",
              "Metrics                                                          \n",
              "Silhouette Coeff                   0.491297           0.491297   \n",
              "Calinski and Harabasz score      255.798088         255.798088   \n",
              "Davies-Bouldin score               0.864104           0.864104   \n",
              "Random index                       0.866621           0.866621   \n",
              "Fowlkes-Mallows score              0.800837           0.800837   \n",
              "\n",
              "                             cosine: k-medoids++  cosine: build  \\\n",
              "Metrics                                                           \n",
              "Silhouette Coeff                        0.491297       0.491297   \n",
              "Calinski and Harabasz score           255.798088     255.798088   \n",
              "Davies-Bouldin score                    0.864104       0.864104   \n",
              "Random index                            0.866621       0.866621   \n",
              "Fowlkes-Mallows score                   0.800837       0.800837   \n",
              "\n",
              "                             manhattan: random  manhattan: heuristic  \\\n",
              "Metrics                                                                \n",
              "Silhouette Coeff                      0.480470              0.480470   \n",
              "Calinski and Harabasz score         243.876470            243.876470   \n",
              "Davies-Bouldin score                  0.894133              0.894133   \n",
              "Random index                          0.876373              0.876373   \n",
              "Fowlkes-Mallows score                 0.814742              0.814742   \n",
              "\n",
              "                             manhattan: k-medoids++  manhattan: build  \n",
              "Metrics                                                                \n",
              "Silhouette Coeff                           0.480470          0.480470  \n",
              "Calinski and Harabasz score              243.876470        243.876470  \n",
              "Davies-Bouldin score                       0.894133          0.894133  \n",
              "Random index                               0.876373          0.876373  \n",
              "Fowlkes-Mallows score                      0.814742          0.814742  "
            ],
            "text/html": [
              "\n",
              "  <div id=\"df-6fe5179c-21df-447d-a47a-808daebedf9d\">\n",
              "    <div class=\"colab-df-container\">\n",
              "      <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>euclidean: random</th>\n",
              "      <th>euclidean: heuristic</th>\n",
              "      <th>euclidean: k-medoids++</th>\n",
              "      <th>euclidean: build</th>\n",
              "      <th>cosine: random</th>\n",
              "      <th>cosine: heuristic</th>\n",
              "      <th>cosine: k-medoids++</th>\n",
              "      <th>cosine: build</th>\n",
              "      <th>manhattan: random</th>\n",
              "      <th>manhattan: heuristic</th>\n",
              "      <th>manhattan: k-medoids++</th>\n",
              "      <th>manhattan: build</th>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>Metrics</th>\n",
              "      <th></th>\n",
              "      <th></th>\n",
              "      <th></th>\n",
              "      <th></th>\n",
              "      <th></th>\n",
              "      <th></th>\n",
              "      <th></th>\n",
              "      <th></th>\n",
              "      <th></th>\n",
              "      <th></th>\n",
              "      <th></th>\n",
              "      <th></th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>Silhouette Coeff</th>\n",
              "      <td>0.493491</td>\n",
              "      <td>0.493491</td>\n",
              "      <td>0.493491</td>\n",
              "      <td>0.493491</td>\n",
              "      <td>0.491297</td>\n",
              "      <td>0.491297</td>\n",
              "      <td>0.491297</td>\n",
              "      <td>0.491297</td>\n",
              "      <td>0.480470</td>\n",
              "      <td>0.480470</td>\n",
              "      <td>0.480470</td>\n",
              "      <td>0.480470</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>Calinski and Harabasz score</th>\n",
              "      <td>258.538210</td>\n",
              "      <td>258.538210</td>\n",
              "      <td>258.538210</td>\n",
              "      <td>258.538210</td>\n",
              "      <td>255.798088</td>\n",
              "      <td>255.798088</td>\n",
              "      <td>255.798088</td>\n",
              "      <td>255.798088</td>\n",
              "      <td>243.876470</td>\n",
              "      <td>243.876470</td>\n",
              "      <td>243.876470</td>\n",
              "      <td>243.876470</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>Davies-Bouldin score</th>\n",
              "      <td>0.854915</td>\n",
              "      <td>0.854915</td>\n",
              "      <td>0.854915</td>\n",
              "      <td>0.854915</td>\n",
              "      <td>0.864104</td>\n",
              "      <td>0.864104</td>\n",
              "      <td>0.864104</td>\n",
              "      <td>0.864104</td>\n",
              "      <td>0.894133</td>\n",
              "      <td>0.894133</td>\n",
              "      <td>0.894133</td>\n",
              "      <td>0.894133</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>Random index</th>\n",
              "      <td>0.861609</td>\n",
              "      <td>0.861609</td>\n",
              "      <td>0.861609</td>\n",
              "      <td>0.861609</td>\n",
              "      <td>0.866621</td>\n",
              "      <td>0.866621</td>\n",
              "      <td>0.866621</td>\n",
              "      <td>0.866621</td>\n",
              "      <td>0.876373</td>\n",
              "      <td>0.876373</td>\n",
              "      <td>0.876373</td>\n",
              "      <td>0.876373</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>Fowlkes-Mallows score</th>\n",
              "      <td>0.792906</td>\n",
              "      <td>0.792906</td>\n",
              "      <td>0.792906</td>\n",
              "      <td>0.792906</td>\n",
              "      <td>0.800837</td>\n",
              "      <td>0.800837</td>\n",
              "      <td>0.800837</td>\n",
              "      <td>0.800837</td>\n",
              "      <td>0.814742</td>\n",
              "      <td>0.814742</td>\n",
              "      <td>0.814742</td>\n",
              "      <td>0.814742</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>\n",
              "      <button class=\"colab-df-convert\" onclick=\"convertToInteractive('df-6fe5179c-21df-447d-a47a-808daebedf9d')\"\n",
              "              title=\"Convert this dataframe to an interactive table.\"\n",
              "              style=\"display:none;\">\n",
              "        \n",
              "  <svg xmlns=\"http://www.w3.org/2000/svg\" height=\"24px\"viewBox=\"0 0 24 24\"\n",
              "       width=\"24px\">\n",
              "    <path d=\"M0 0h24v24H0V0z\" fill=\"none\"/>\n",
              "    <path d=\"M18.56 5.44l.94 2.06.94-2.06 2.06-.94-2.06-.94-.94-2.06-.94 2.06-2.06.94zm-11 1L8.5 8.5l.94-2.06 2.06-.94-2.06-.94L8.5 2.5l-.94 2.06-2.06.94zm10 10l.94 2.06.94-2.06 2.06-.94-2.06-.94-.94-2.06-.94 2.06-2.06.94z\"/><path d=\"M17.41 7.96l-1.37-1.37c-.4-.4-.92-.59-1.43-.59-.52 0-1.04.2-1.43.59L10.3 9.45l-7.72 7.72c-.78.78-.78 2.05 0 2.83L4 21.41c.39.39.9.59 1.41.59.51 0 1.02-.2 1.41-.59l7.78-7.78 2.81-2.81c.8-.78.8-2.07 0-2.86zM5.41 20L4 18.59l7.72-7.72 1.47 1.35L5.41 20z\"/>\n",
              "  </svg>\n",
              "      </button>\n",
              "      \n",
              "  <style>\n",
              "    .colab-df-container {\n",
              "      display:flex;\n",
              "      flex-wrap:wrap;\n",
              "      gap: 12px;\n",
              "    }\n",
              "\n",
              "    .colab-df-convert {\n",
              "      background-color: #E8F0FE;\n",
              "      border: none;\n",
              "      border-radius: 50%;\n",
              "      cursor: pointer;\n",
              "      display: none;\n",
              "      fill: #1967D2;\n",
              "      height: 32px;\n",
              "      padding: 0 0 0 0;\n",
              "      width: 32px;\n",
              "    }\n",
              "\n",
              "    .colab-df-convert:hover {\n",
              "      background-color: #E2EBFA;\n",
              "      box-shadow: 0px 1px 2px rgba(60, 64, 67, 0.3), 0px 1px 3px 1px rgba(60, 64, 67, 0.15);\n",
              "      fill: #174EA6;\n",
              "    }\n",
              "\n",
              "    [theme=dark] .colab-df-convert {\n",
              "      background-color: #3B4455;\n",
              "      fill: #D2E3FC;\n",
              "    }\n",
              "\n",
              "    [theme=dark] .colab-df-convert:hover {\n",
              "      background-color: #434B5C;\n",
              "      box-shadow: 0px 1px 3px 1px rgba(0, 0, 0, 0.15);\n",
              "      filter: drop-shadow(0px 1px 2px rgba(0, 0, 0, 0.3));\n",
              "      fill: #FFFFFF;\n",
              "    }\n",
              "  </style>\n",
              "\n",
              "      <script>\n",
              "        const buttonEl =\n",
              "          document.querySelector('#df-6fe5179c-21df-447d-a47a-808daebedf9d button.colab-df-convert');\n",
              "        buttonEl.style.display =\n",
              "          google.colab.kernel.accessAllowed ? 'block' : 'none';\n",
              "\n",
              "        async function convertToInteractive(key) {\n",
              "          const element = document.querySelector('#df-6fe5179c-21df-447d-a47a-808daebedf9d');\n",
              "          const dataTable =\n",
              "            await google.colab.kernel.invokeFunction('convertToInteractive',\n",
              "                                                     [key], {});\n",
              "          if (!dataTable) return;\n",
              "\n",
              "          const docLinkHtml = 'Like what you see? Visit the ' +\n",
              "            '<a target=\"_blank\" href=https://colab.research.google.com/notebooks/data_table.ipynb>data table notebook</a>'\n",
              "            + ' to learn more about interactive tables.';\n",
              "          element.innerHTML = '';\n",
              "          dataTable['output_type'] = 'display_data';\n",
              "          await google.colab.output.renderOutput(dataTable, element);\n",
              "          const docLink = document.createElement('div');\n",
              "          docLink.innerHTML = docLinkHtml;\n",
              "          element.appendChild(docLink);\n",
              "        }\n",
              "      </script>\n",
              "    </div>\n",
              "  </div>\n",
              "  "
            ]
          },
          "metadata": {},
          "execution_count": 12
        }
      ],
      "source": [
        "#creating a dataframe from dictionary to visualise the results\n",
        "df_seed=pd.DataFrame(dict_metrics_seed)\n",
        "#setting index of the datframe to Metrics\n",
        "df_seed.set_index('Metrics',inplace=True)\n",
        "#Presenting the results in a tabular form\n",
        "print('Presenting the results in a tabular form')\n",
        "df_seed"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "8ObBYlmJ-1V8"
      },
      "source": [
        "### **Observations from  Kmedoid hyperparameter tuning on the dataset**\n",
        "\n",
        "* When we fix metric and vary init: \n",
        "\n",
        "  * It can be observed that once the *metric* type parameter is fixed then the variation of *init* parameter has no impact for any given cluster validation index.\n",
        "\n",
        "\n",
        "* when init is fixed (k-medoids++) and we vary *metric* :\n",
        "  * For extrinsic measures the performance of various metric is in order \n",
        "    * manhattan> cosine> euclidean\n",
        "\n",
        "  * For intrinsic measures the performance of various metric is in order:\n",
        "    * euclidean> cosine > manhattan\n",
        "\n",
        "\n",
        "\n",
        "\n",
        "\n",
        "\n",
        "\n"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 255
        },
        "id": "3sa_MMlV73_t",
        "outputId": "7edba6d6-f915-4506-9d92-980e2f3036eb"
      },
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "fixing the init and varying along the metric \n"
          ]
        },
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "                             euclidean: k-medoids++  cosine: k-medoids++  \\\n",
              "Metrics                                                                    \n",
              "Silhouette Coeff                           0.493491             0.491297   \n",
              "Calinski and Harabasz score              258.538210           255.798088   \n",
              "Davies-Bouldin score                       0.854915             0.864104   \n",
              "Random index                               0.861609             0.866621   \n",
              "Fowlkes-Mallows score                      0.792906             0.800837   \n",
              "\n",
              "                             manhattan: k-medoids++  \n",
              "Metrics                                              \n",
              "Silhouette Coeff                           0.480470  \n",
              "Calinski and Harabasz score              243.876470  \n",
              "Davies-Bouldin score                       0.894133  \n",
              "Random index                               0.876373  \n",
              "Fowlkes-Mallows score                      0.814742  "
            ],
            "text/html": [
              "\n",
              "  <div id=\"df-1452fc9c-a3c7-4843-a3a1-dbbf0d09fd81\">\n",
              "    <div class=\"colab-df-container\">\n",
              "      <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>euclidean: k-medoids++</th>\n",
              "      <th>cosine: k-medoids++</th>\n",
              "      <th>manhattan: k-medoids++</th>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>Metrics</th>\n",
              "      <th></th>\n",
              "      <th></th>\n",
              "      <th></th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>Silhouette Coeff</th>\n",
              "      <td>0.493491</td>\n",
              "      <td>0.491297</td>\n",
              "      <td>0.480470</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>Calinski and Harabasz score</th>\n",
              "      <td>258.538210</td>\n",
              "      <td>255.798088</td>\n",
              "      <td>243.876470</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>Davies-Bouldin score</th>\n",
              "      <td>0.854915</td>\n",
              "      <td>0.864104</td>\n",
              "      <td>0.894133</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>Random index</th>\n",
              "      <td>0.861609</td>\n",
              "      <td>0.866621</td>\n",
              "      <td>0.876373</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>Fowlkes-Mallows score</th>\n",
              "      <td>0.792906</td>\n",
              "      <td>0.800837</td>\n",
              "      <td>0.814742</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>\n",
              "      <button class=\"colab-df-convert\" onclick=\"convertToInteractive('df-1452fc9c-a3c7-4843-a3a1-dbbf0d09fd81')\"\n",
              "              title=\"Convert this dataframe to an interactive table.\"\n",
              "              style=\"display:none;\">\n",
              "        \n",
              "  <svg xmlns=\"http://www.w3.org/2000/svg\" height=\"24px\"viewBox=\"0 0 24 24\"\n",
              "       width=\"24px\">\n",
              "    <path d=\"M0 0h24v24H0V0z\" fill=\"none\"/>\n",
              "    <path d=\"M18.56 5.44l.94 2.06.94-2.06 2.06-.94-2.06-.94-.94-2.06-.94 2.06-2.06.94zm-11 1L8.5 8.5l.94-2.06 2.06-.94-2.06-.94L8.5 2.5l-.94 2.06-2.06.94zm10 10l.94 2.06.94-2.06 2.06-.94-2.06-.94-.94-2.06-.94 2.06-2.06.94z\"/><path d=\"M17.41 7.96l-1.37-1.37c-.4-.4-.92-.59-1.43-.59-.52 0-1.04.2-1.43.59L10.3 9.45l-7.72 7.72c-.78.78-.78 2.05 0 2.83L4 21.41c.39.39.9.59 1.41.59.51 0 1.02-.2 1.41-.59l7.78-7.78 2.81-2.81c.8-.78.8-2.07 0-2.86zM5.41 20L4 18.59l7.72-7.72 1.47 1.35L5.41 20z\"/>\n",
              "  </svg>\n",
              "      </button>\n",
              "      \n",
              "  <style>\n",
              "    .colab-df-container {\n",
              "      display:flex;\n",
              "      flex-wrap:wrap;\n",
              "      gap: 12px;\n",
              "    }\n",
              "\n",
              "    .colab-df-convert {\n",
              "      background-color: #E8F0FE;\n",
              "      border: none;\n",
              "      border-radius: 50%;\n",
              "      cursor: pointer;\n",
              "      display: none;\n",
              "      fill: #1967D2;\n",
              "      height: 32px;\n",
              "      padding: 0 0 0 0;\n",
              "      width: 32px;\n",
              "    }\n",
              "\n",
              "    .colab-df-convert:hover {\n",
              "      background-color: #E2EBFA;\n",
              "      box-shadow: 0px 1px 2px rgba(60, 64, 67, 0.3), 0px 1px 3px 1px rgba(60, 64, 67, 0.15);\n",
              "      fill: #174EA6;\n",
              "    }\n",
              "\n",
              "    [theme=dark] .colab-df-convert {\n",
              "      background-color: #3B4455;\n",
              "      fill: #D2E3FC;\n",
              "    }\n",
              "\n",
              "    [theme=dark] .colab-df-convert:hover {\n",
              "      background-color: #434B5C;\n",
              "      box-shadow: 0px 1px 3px 1px rgba(0, 0, 0, 0.15);\n",
              "      filter: drop-shadow(0px 1px 2px rgba(0, 0, 0, 0.3));\n",
              "      fill: #FFFFFF;\n",
              "    }\n",
              "  </style>\n",
              "\n",
              "      <script>\n",
              "        const buttonEl =\n",
              "          document.querySelector('#df-1452fc9c-a3c7-4843-a3a1-dbbf0d09fd81 button.colab-df-convert');\n",
              "        buttonEl.style.display =\n",
              "          google.colab.kernel.accessAllowed ? 'block' : 'none';\n",
              "\n",
              "        async function convertToInteractive(key) {\n",
              "          const element = document.querySelector('#df-1452fc9c-a3c7-4843-a3a1-dbbf0d09fd81');\n",
              "          const dataTable =\n",
              "            await google.colab.kernel.invokeFunction('convertToInteractive',\n",
              "                                                     [key], {});\n",
              "          if (!dataTable) return;\n",
              "\n",
              "          const docLinkHtml = 'Like what you see? Visit the ' +\n",
              "            '<a target=\"_blank\" href=https://colab.research.google.com/notebooks/data_table.ipynb>data table notebook</a>'\n",
              "            + ' to learn more about interactive tables.';\n",
              "          element.innerHTML = '';\n",
              "          dataTable['output_type'] = 'display_data';\n",
              "          await google.colab.output.renderOutput(dataTable, element);\n",
              "          const docLink = document.createElement('div');\n",
              "          docLink.innerHTML = docLinkHtml;\n",
              "          element.appendChild(docLink);\n",
              "        }\n",
              "      </script>\n",
              "    </div>\n",
              "  </div>\n",
              "  "
            ]
          },
          "metadata": {},
          "execution_count": 13
        }
      ],
      "source": [
        "# displaying the dataframe by fixing the init and varying along the metric \n",
        "print('fixing the init and varying along the metric ')\n",
        "df_seed.iloc[:,[2,6,10]]"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "QcbvRkJGAYPk"
      },
      "source": [
        "## **Conclusion**\n",
        "Looking at the above dataframe we select manhattan  metric and k-medoids++ as our choice  of  \n",
        "metric hyperparameter value"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "w2R56Kadmp2a"
      },
      "source": [
        "## **Hyperparameter tuning CLARANS**"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "xwQz_oUDmqnK",
        "outputId": "b46a19e1-dd50-4cd5-ef5b-7270700f143b"
      },
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "****************************\n",
            "numlocal: 2; maxneighbor: 2\n",
            "Intrinsic Measure:\n",
            "\n",
            "Silhouette Coefficient: 0.485\n",
            "Calinski and Harabasz score: 249.661\n",
            "Davies-Bouldin score: 0.866\n",
            "\n",
            "Extrinsic Measure:\n",
            "\n",
            "Random index: 0.854\n",
            "Fowlkes-Mallows score: 0.784\n",
            "\n",
            "****************************\n",
            "numlocal: 2; maxneighbor: 3\n",
            "Intrinsic Measure:\n",
            "\n",
            "Silhouette Coefficient: 0.426\n",
            "Calinski and Harabasz score: 176.149\n",
            "Davies-Bouldin score: 0.967\n",
            "\n",
            "Extrinsic Measure:\n",
            "\n",
            "Random index: 0.845\n",
            "Fowlkes-Mallows score: 0.768\n",
            "\n",
            "****************************\n",
            "numlocal: 3; maxneighbor: 2\n",
            "Intrinsic Measure:\n",
            "\n",
            "Silhouette Coefficient: 0.467\n",
            "Calinski and Harabasz score: 230.343\n",
            "Davies-Bouldin score: 0.892\n",
            "\n",
            "Extrinsic Measure:\n",
            "\n",
            "Random index: 0.829\n",
            "Fowlkes-Mallows score: 0.755\n",
            "\n",
            "****************************\n",
            "numlocal: 3; maxneighbor: 3\n",
            "Intrinsic Measure:\n",
            "\n",
            "Silhouette Coefficient: 0.470\n",
            "Calinski and Harabasz score: 233.230\n",
            "Davies-Bouldin score: 0.875\n",
            "\n",
            "Extrinsic Measure:\n",
            "\n",
            "Random index: 0.834\n",
            "Fowlkes-Mallows score: 0.761\n",
            "\n",
            "****************************\n",
            "numlocal: 4; maxneighbor: 2\n",
            "Intrinsic Measure:\n",
            "\n",
            "Silhouette Coefficient: 0.493\n",
            "Calinski and Harabasz score: 258.110\n",
            "Davies-Bouldin score: 0.856\n",
            "\n",
            "Extrinsic Measure:\n",
            "\n",
            "Random index: 0.857\n",
            "Fowlkes-Mallows score: 0.786\n",
            "\n",
            "****************************\n",
            "numlocal: 4; maxneighbor: 3\n",
            "Intrinsic Measure:\n",
            "\n",
            "Silhouette Coefficient: 0.406\n",
            "Calinski and Harabasz score: 185.914\n",
            "Davies-Bouldin score: 0.987\n",
            "\n",
            "Extrinsic Measure:\n",
            "\n",
            "Random index: 0.792\n",
            "Fowlkes-Mallows score: 0.703\n",
            "\n"
          ]
        }
      ],
      "source": [
        "# Clustering using ClARANS\n",
        "\n",
        "#high computing resources are required for  running CALARNS\n",
        "\n",
        "#so only lower values of hyperparameters have been tested in the absence of high computing values\n",
        "\n",
        "\n",
        "dict_metrics_seedclara={}\n",
        "dict_metrics_seedclara['Metrics']=['Silhouette Coeff','Calinski and Harabasz score','Davies-Bouldin score','Random index','Fowlkes-Mallows score']\n",
        "\n",
        "local_min=[2,3,4] # various values of local minima \n",
        "neighbor=[2,3] # # various values of max neighbor\n",
        "\n",
        "\n",
        "for numlocal in local_min:\n",
        "  for maxneighbor in neighbor:\n",
        "\n",
        "    seed_copy=seed.copy()\n",
        "    d={}\n",
        "    clarans_instance = clarans(normalized_df_seed, number_clusters=3, numlocal=numlocal, maxneighbor=maxneighbor)\n",
        "    clarans_instance.process()\n",
        "    d['clusters'] = clarans_instance.get_clusters()\n",
        "    d['medoids'] = clarans_instance.get_medoids()\n",
        "\n",
        "    indices_0=d['clusters'][0]\n",
        "    indices_1=d['clusters'][1]\n",
        "    indices_2=d['clusters'][2]\n",
        "    seed_copy['CLARANS cluster labels']=[-1]* len(seed_copy)\n",
        "\n",
        "    seed_copy.iloc[seed_copy.index.isin(indices_0),[8]]=0\n",
        "    seed_copy.iloc[seed_copy.index.isin(indices_1),[8]]=1\n",
        "    seed_copy.iloc[seed_copy.index.isin(indices_2),[8]]=2\n",
        "\n",
        "\n",
        "\n",
        "    y_true=seed_labels.tolist()\n",
        "    y_true=np.array(y_true).reshape(-1)\n",
        "    \n",
        "    y_pred=seed_copy['CLARANS cluster labels'].values.tolist()\n",
        "    SIL_ward=metrics.silhouette_score(normalized_df_seed, y_pred)\n",
        "    CH_ward=metrics.calinski_harabasz_score(normalized_df_seed, y_pred)\n",
        "    DB_ward=metrics.davies_bouldin_score( normalized_df_seed, y_pred)\n",
        "    \n",
        "    RI_ward=metrics.rand_score(y_true,y_pred)\n",
        "    FM_ward=metrics.fowlkes_mallows_score(y_true,y_pred)\n",
        "\n",
        "    dict_metrics_seedclara['local-'+str(numlocal)+str(';')+'neighbor-'+str(maxneighbor)]=[SIL_ward,CH_ward,DB_ward,RI_ward,FM_ward]\n",
        "\n",
        "    print('****************************')\n",
        "    print(f'numlocal: {numlocal}; maxneighbor: {maxneighbor}')\n",
        "    print('Intrinsic Measure:\\n')\n",
        "    print(f'Silhouette Coefficient: {SIL_ward:.3f}')\n",
        "    print(f'Calinski and Harabasz score: {CH_ward:.3f}')\n",
        "    print(f'Davies-Bouldin score: {DB_ward:.3f}')\n",
        "    print('')\n",
        "    print('Extrinsic Measure:\\n')\n",
        "    print(f'Random index: {RI_ward:.3f}')\n",
        "    print(f'Fowlkes-Mallows score: {FM_ward:.3f}')\n",
        "\n",
        "    \n",
        "    print()\n",
        "\n",
        "\n"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 206
        },
        "id": "kQ6fLRZi7l_A",
        "outputId": "d434b4a8-7a06-4c45-8b66-d17bd6ec0e50"
      },
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "                       Metrics  local-2;neighbor-2  local-2;neighbor-3  \\\n",
              "0             Silhouette Coeff            0.485331            0.425599   \n",
              "1  Calinski and Harabasz score          249.660873          176.149069   \n",
              "2         Davies-Bouldin score            0.866476            0.966811   \n",
              "3                 Random index            0.853862            0.844520   \n",
              "4        Fowlkes-Mallows score            0.784129            0.767738   \n",
              "\n",
              "   local-3;neighbor-2  local-3;neighbor-3  local-4;neighbor-2  \\\n",
              "0            0.466842            0.469921            0.492842   \n",
              "1          230.343290          233.230318          258.110411   \n",
              "2            0.891979            0.874978            0.855619   \n",
              "3            0.829255            0.834450            0.856869   \n",
              "4            0.754513            0.761082            0.786147   \n",
              "\n",
              "   local-4;neighbor-3  \n",
              "0            0.406371  \n",
              "1          185.914208  \n",
              "2            0.987046  \n",
              "3            0.792390  \n",
              "4            0.703078  "
            ],
            "text/html": [
              "\n",
              "  <div id=\"df-478073f5-588b-4ab3-a58b-cf6c19816f13\">\n",
              "    <div class=\"colab-df-container\">\n",
              "      <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>Metrics</th>\n",
              "      <th>local-2;neighbor-2</th>\n",
              "      <th>local-2;neighbor-3</th>\n",
              "      <th>local-3;neighbor-2</th>\n",
              "      <th>local-3;neighbor-3</th>\n",
              "      <th>local-4;neighbor-2</th>\n",
              "      <th>local-4;neighbor-3</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>Silhouette Coeff</td>\n",
              "      <td>0.485331</td>\n",
              "      <td>0.425599</td>\n",
              "      <td>0.466842</td>\n",
              "      <td>0.469921</td>\n",
              "      <td>0.492842</td>\n",
              "      <td>0.406371</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1</th>\n",
              "      <td>Calinski and Harabasz score</td>\n",
              "      <td>249.660873</td>\n",
              "      <td>176.149069</td>\n",
              "      <td>230.343290</td>\n",
              "      <td>233.230318</td>\n",
              "      <td>258.110411</td>\n",
              "      <td>185.914208</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2</th>\n",
              "      <td>Davies-Bouldin score</td>\n",
              "      <td>0.866476</td>\n",
              "      <td>0.966811</td>\n",
              "      <td>0.891979</td>\n",
              "      <td>0.874978</td>\n",
              "      <td>0.855619</td>\n",
              "      <td>0.987046</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3</th>\n",
              "      <td>Random index</td>\n",
              "      <td>0.853862</td>\n",
              "      <td>0.844520</td>\n",
              "      <td>0.829255</td>\n",
              "      <td>0.834450</td>\n",
              "      <td>0.856869</td>\n",
              "      <td>0.792390</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4</th>\n",
              "      <td>Fowlkes-Mallows score</td>\n",
              "      <td>0.784129</td>\n",
              "      <td>0.767738</td>\n",
              "      <td>0.754513</td>\n",
              "      <td>0.761082</td>\n",
              "      <td>0.786147</td>\n",
              "      <td>0.703078</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>\n",
              "      <button class=\"colab-df-convert\" onclick=\"convertToInteractive('df-478073f5-588b-4ab3-a58b-cf6c19816f13')\"\n",
              "              title=\"Convert this dataframe to an interactive table.\"\n",
              "              style=\"display:none;\">\n",
              "        \n",
              "  <svg xmlns=\"http://www.w3.org/2000/svg\" height=\"24px\"viewBox=\"0 0 24 24\"\n",
              "       width=\"24px\">\n",
              "    <path d=\"M0 0h24v24H0V0z\" fill=\"none\"/>\n",
              "    <path d=\"M18.56 5.44l.94 2.06.94-2.06 2.06-.94-2.06-.94-.94-2.06-.94 2.06-2.06.94zm-11 1L8.5 8.5l.94-2.06 2.06-.94-2.06-.94L8.5 2.5l-.94 2.06-2.06.94zm10 10l.94 2.06.94-2.06 2.06-.94-2.06-.94-.94-2.06-.94 2.06-2.06.94z\"/><path d=\"M17.41 7.96l-1.37-1.37c-.4-.4-.92-.59-1.43-.59-.52 0-1.04.2-1.43.59L10.3 9.45l-7.72 7.72c-.78.78-.78 2.05 0 2.83L4 21.41c.39.39.9.59 1.41.59.51 0 1.02-.2 1.41-.59l7.78-7.78 2.81-2.81c.8-.78.8-2.07 0-2.86zM5.41 20L4 18.59l7.72-7.72 1.47 1.35L5.41 20z\"/>\n",
              "  </svg>\n",
              "      </button>\n",
              "      \n",
              "  <style>\n",
              "    .colab-df-container {\n",
              "      display:flex;\n",
              "      flex-wrap:wrap;\n",
              "      gap: 12px;\n",
              "    }\n",
              "\n",
              "    .colab-df-convert {\n",
              "      background-color: #E8F0FE;\n",
              "      border: none;\n",
              "      border-radius: 50%;\n",
              "      cursor: pointer;\n",
              "      display: none;\n",
              "      fill: #1967D2;\n",
              "      height: 32px;\n",
              "      padding: 0 0 0 0;\n",
              "      width: 32px;\n",
              "    }\n",
              "\n",
              "    .colab-df-convert:hover {\n",
              "      background-color: #E2EBFA;\n",
              "      box-shadow: 0px 1px 2px rgba(60, 64, 67, 0.3), 0px 1px 3px 1px rgba(60, 64, 67, 0.15);\n",
              "      fill: #174EA6;\n",
              "    }\n",
              "\n",
              "    [theme=dark] .colab-df-convert {\n",
              "      background-color: #3B4455;\n",
              "      fill: #D2E3FC;\n",
              "    }\n",
              "\n",
              "    [theme=dark] .colab-df-convert:hover {\n",
              "      background-color: #434B5C;\n",
              "      box-shadow: 0px 1px 3px 1px rgba(0, 0, 0, 0.15);\n",
              "      filter: drop-shadow(0px 1px 2px rgba(0, 0, 0, 0.3));\n",
              "      fill: #FFFFFF;\n",
              "    }\n",
              "  </style>\n",
              "\n",
              "      <script>\n",
              "        const buttonEl =\n",
              "          document.querySelector('#df-478073f5-588b-4ab3-a58b-cf6c19816f13 button.colab-df-convert');\n",
              "        buttonEl.style.display =\n",
              "          google.colab.kernel.accessAllowed ? 'block' : 'none';\n",
              "\n",
              "        async function convertToInteractive(key) {\n",
              "          const element = document.querySelector('#df-478073f5-588b-4ab3-a58b-cf6c19816f13');\n",
              "          const dataTable =\n",
              "            await google.colab.kernel.invokeFunction('convertToInteractive',\n",
              "                                                     [key], {});\n",
              "          if (!dataTable) return;\n",
              "\n",
              "          const docLinkHtml = 'Like what you see? Visit the ' +\n",
              "            '<a target=\"_blank\" href=https://colab.research.google.com/notebooks/data_table.ipynb>data table notebook</a>'\n",
              "            + ' to learn more about interactive tables.';\n",
              "          element.innerHTML = '';\n",
              "          dataTable['output_type'] = 'display_data';\n",
              "          await google.colab.output.renderOutput(dataTable, element);\n",
              "          const docLink = document.createElement('div');\n",
              "          docLink.innerHTML = docLinkHtml;\n",
              "          element.appendChild(docLink);\n",
              "        }\n",
              "      </script>\n",
              "    </div>\n",
              "  </div>\n",
              "  "
            ]
          },
          "metadata": {},
          "execution_count": 15
        }
      ],
      "source": [
        "df3_clara=pd.DataFrame(dict_metrics_seedclara)\n",
        "df3_clara"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "vcaANR_NYRsl"
      },
      "source": [
        "### **Observations from CLARANS hyperparameter tuning**\n",
        "* The best results cane be seen for the values:\n",
        "  * numlocal: 4\n",
        "  * maxneighbor: 2"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "p6ee9eiQIqbj"
      },
      "source": [
        "# References\n",
        "* https://scikit-learn-extra.readthedocs.io/en/stable/generated/sklearn_extra.cluster.KMedoids.html\n",
        "* Park, H.S.and Jun, C.H., 2009. A simple and fast algorithm for K-medoids clustering.  Expert systems with applications, 36(2), pp.3336-3341\n",
        "    https://www.sciencedirect.com/science/article/pii/S095741740800081X?via%3Dihub\n"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "6wQxCztbIsLt"
      },
      "outputs": [],
      "source": []
    }
  ],
  "metadata": {
    "colab": {
      "provenance": [],
      "toc_visible": true
    },
    "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": "3.9.12"
    }
  },
  "nbformat": 4,
  "nbformat_minor": 0
}