slt-ce-5.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {
    "colab_type": "text",
    "id": "cr1DzovqNHd3"
   },
   "source": [
    "# SLT-CE-5: Pairwise Clustering"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "#"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "colab_type": "text",
    "id": "QT4CPEmgNHd6"
   },
   "source": [
    "In this coding exercise you will apply pairwise clustering for proximity data and you will see how to use the mean-field approximation (MFA) approach with the EM-algorithm to get a labeling for a data object encoded by similarity and dissimilarity matrices. You will also be able to implement modifications of the EM-scheme to obtain embeddings into an Euclidean space.\n",
    "\n",
    "<b> References: </b>\n",
    "\n",
    "[1] T. Hoffman, JM Buhmann. Pairwise Data Clustering by Deterministic Annealing, 1996: http://citeseerx.ist.psu.edu/viewdoc/download?doi=10.1.1.8.9812&rep=rep1&type=pdf \n",
    "\n",
    "Optionally, the following resources could be helpful:\n",
    "\n",
    "[2] Lecture 7. Pairwise clustering: https://ml2.inf.ethz.ch/courses/slt/lectures/slt20_lecture07.pdf\n",
    "\n",
    "[3] Lecture script. Chapter 3.2: https://ml2.inf.ethz.ch/courses/slt/tutorials/SLT_Script.pdf\n",
    "\n",
    "[4] Lecture 4. Determenistic Annealing: https://ml2.inf.ethz.ch/courses/slt/lectures/slt20_lecture04.pdf\n",
    "\n",
    "[5] Lecture 8. Mean-field approximation: https://ml2.inf.ethz.ch/courses/slt/lectures/slt20_lecture08.pdf"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "colab_type": "text",
    "id": "-9PbKO3KNHd7"
   },
   "source": [
    "<h2 style=\"background-color:#f0b375;\"> Setup </h2>"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {
    "colab": {},
    "colab_type": "code",
    "id": "9ZGkoTLsNHd8"
   },
   "outputs": [],
   "source": [
    "import sklearn as skl\n",
    "import matplotlib.pyplot as plt\n",
    "import numpy as np\n",
    "import scipy as sp\n",
    "import pandas as pd\n",
    "import sys\n",
    "from scipy.stats import mode\n",
    "import warnings\n",
    "warnings.filterwarnings('ignore')\n",
    "\n",
    "from sklearn.utils.validation import check_is_fitted\n",
    "from sklearn.cluster import SpectralClustering\n",
    "from scipy.io import loadmat\n",
    "\n",
    "#Please use this seed in all sections, except section 5.5\n",
    "np.random.seed(111)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "colab_type": "text",
    "id": "zRklPbiHNHeC"
   },
   "source": [
    "<h2 style=\"background-color:#f0b375;\"> Data </h2>"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "colab_type": "text",
    "id": "sOOSq-gMNHeD"
   },
   "source": [
    "## ProDom Data\n",
    "\n",
    "ProDom is a comprehensive set of protein domain families [6]. A ProDom subset of 2604 protein domain sequences from the ProDom set was selected by [7]. These are chosen based on a high similarity to at least one sequence contained in the first four folds of the SCOP database. The pairwise structural alignments are computed [7]. Each SCOP sequence belongs to a group, as labeled by the experts [8]. The same four classes are assigned here.\n",
    "\n",
    "<b> References: </b>\n",
    "\n",
    "[6] F. Corpet, F. Servant, J. Gouzy and D. Kahn, ProDom and ProDom-CG: tools for protein domain analysis and whole genome comparisons, Nucleid Acids Res., vol. 28, 267-269, 2000.\n",
    "\n",
    "[7] V. Roth, J. Laub, J.M. Buhmann, and K.-R. Mueller, Going metric: Denoising pairwise data, Advances in Neural Information Processing Systems, 841-856, MIT Press, 2003.\n",
    "\n",
    "[8] A.G. Murzin, S.E. Brenner, T. Hubbard and C. Chothia, SCOP: a structural classification of proteins database for the investigation of sequences and structures, Journal of Molecular Biology, vol. 247, 536-540, 1995.\n",
    "\n",
    "\n",
    "Download link: http://prtools.org/disdatasets/index.html#ProDom"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {
    "colab": {},
    "colab_type": "code",
    "id": "ooEavYIkNHeE"
   },
   "outputs": [],
   "source": [
    "prodom = {\"input\": loadmat(\"prodom.mat\")}\n",
    "# After some reverse-engineering:\n",
    "prodom[\"similarity\"] = np.array(prodom[\"input\"][\"s\"])[0,0][0]\n",
    "prodom[\"labels\"] =np.array(prodom[\"input\"][\"s\"])[0,0][2].flatten()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "colab_type": "text",
    "id": "UY5KO2PMNHeH"
   },
   "source": [
    "## Protein Data\n",
    "\n",
    "The protein data are provided as a 213x213 dissimilarity matrix comparing the protein sequences based on the concept of an evolutionary distance. It was used for classification in [9] and for clustering in [10]. There are four classes of globins: heterogeneous globin (G), hemoglobin-A (HA), hemoglobin-B (HB) and myoglobin (M). Each protein in the dataset is classified into one of these four classes.\n",
    "\n",
    "<b> References: </b>\n",
    "\n",
    "[9] T. Graepel, R. Herbrich, P. Bollmann-Sdorra, K. Obermayer, Classification on pairwise proximity data. In Advances in Neural Information System Processing vol. 11, 438-444, 1999.\n",
    "\n",
    "[10] T. Denoeux, T. and M.-H. Masson, EVCLUS: Evidential clustering of proximity data. IEEE Transations on Systems, Man and Cybernetics, vol. 34, 95-109, 2004.\n",
    "\n",
    "Download link (use The PRTools version of Protein): http://prtools.org/disdatasets/index.html#Protein"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {
    "colab": {},
    "colab_type": "code",
    "id": "VH0cw6_ENHeI"
   },
   "outputs": [],
   "source": [
    "protein = {\"input\": loadmat(\"protein.mat\")}\n",
    "protein[\"dissimilarity\"] = np.array(protein[\"input\"][\"d\"])[0,0][0]\n",
    "protein[\"labels\"] =np.array(protein[\"input\"][\"d\"])[0,0][2].flatten()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "colab_type": "text",
    "id": "5E6Pnq4NNHeM"
   },
   "source": [
    "<h2 style=\"background-color:#f0b375;\">\n",
    "Section 4.0 \n",
    "<span style=font-size:50%> Complete all problems (4.0.1-4.0.9) in this section to get a pass on this exercise. </span>\n",
    "</h2>"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "In this section, you have to give a description of the pairwise clustering approach based on deterministic annealing and mean-field approximation. Then you have to implement it, as described in Algorithm II in reference [1]."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "\n",
    "<p style=\"background-color:#adebad;\"> \n",
    "<b>NOTE</b>: Please follow these guidelines: <br>\n",
    "</p>\n",
    "<ul style=\"background-color:#adebad;\"> \n",
    "<li> Use consistent notation throughout the notebook. </li>\n",
    "<li> Define every variable properly. </li>\n",
    "<li> Define every math notation before usage. </li>\n",
    "<li> Use math mode with $$ for formulas. </li>\n",
    "</ul>\n",
    "\n",
    "\n",
    "<p style=\"background-color:#adebad;\">\n",
    "    1. Implement the function <b>plot_pairwise()</b> according to its docstring. Hint: <b>usematplotlib.pyplot.matshow()</b> to visualize matricies, <b>matplotlib.pyplot.axhspan(),matplotlib.pyplot.axvspan()</b> to add horizontal and vertical rectangles to the plot.\n",
    "</p>"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {
    "colab": {},
    "colab_type": "code",
    "id": "ZNh3zWonNHeN"
   },
   "outputs": [],
   "source": [
    "def plot_pairwise(data, labels, cmap=\"binary\", alpha=0.33):\n",
    "    \"\"\"Plot (dis-)similarity data\n",
    "    \n",
    "    Args:\n",
    "        data (np.ndarray): Symmetric (dis-)similarity array (n_samples, n_samples)\n",
    "        labels (np.array): Cluster label array  (n_samples,)\n",
    "        cmap (str): Color Map for (dis-)similarity values. Please use matplotlib colormaps. See here for details: \n",
    "        https://matplotlib.org/tutorials/colors/colormaps.html\n",
    "        alpha (float): Opacity of cluster overlay, between 0 and 1\n",
    "        \n",
    "    Shows:\n",
    "        Image plot of (dis-)similarity matrix, overlaid with transparent blocks,\n",
    "        which indicate the clusters contained in labels.\n",
    "        Note: Permute the data according to the sorting of the labels.\n",
    "        \n",
    "    Prints:\n",
    "        Minimum, Maximum and Median of (dis-)similarity values\n",
    "        Warning, if (dis-)similarity matrix is not symmetric\n",
    "        \n",
    "    Returns:\n",
    "        This is void function. No return required.\n",
    "    \"\"\"\n",
    "   \n",
    "       # Indexes of the sorted labels (clusters)\n",
    "    idx    = labels.argsort(kind=\"stable\")\n",
    "    labels = labels[idx]\n",
    "    \n",
    "    _, cnt = np.unique(labels, return_counts=True)\n",
    "    util   = np.cumsum(cnt)\n",
    "    util_n = util/util[-1]\n",
    "    \n",
    "    # Permute (dis)similarity matrix according to the labels\n",
    "    P   = data[:, idx]\n",
    "    P   = P[idx,:]\n",
    "    \n",
    "    print(\"data.min = {}\".format(np.min(P)))\n",
    "    print(\"data.median = {}\".format(np.median(P)))\n",
    "    print(\"data.max = {}\".format(np.max(P)))\n",
    "    \n",
    "    # Plot the (dis)similarity matrix\n",
    "    fig, ax = plt.subplots()\n",
    "    ax.matshow(P, cmap=cmap)\n",
    "    ax.axhspan(0, util[0], xmin=0, xmax=util_n[0], facecolor='green', alpha=alpha*0.2)\n",
    "    ax.axhspan(util[0], util[1], xmin=util_n[0], xmax=util_n[1], facecolor='yellow', alpha=alpha*0.2)\n",
    "    ax.axhspan(util[1], util[2], xmin=util_n[1], xmax=util_n[2], facecolor='cyan', alpha=alpha*0.2)\n",
    "    ax.axhspan(util[2], util[3], xmin=util_n[2], xmax=1, facecolor='blue', alpha=alpha*0.2)\n",
    "    \n",
    "    ax.set_ylim(util[-1], 0)\n",
    "    plt.show()\n",
    "    \n",
    "    return"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 320
    },
    "colab_type": "code",
    "id": "hNhpLe4aNHeQ",
    "outputId": "92efe7aa-40e5-48e1-ba2d-6976775a1aa1"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "data.min = -15529\n",
      "data.median = -8847.0\n",
      "data.max = 478835\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "#Example how output should looks like after the following command\n",
    "plot_pairwise(prodom[\"similarity\"], prodom[\"labels\"])"
   ]
  },
  {
   "attachments": {
    "prodom_data_visualization.png": {
     "image/png": "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"
    }
   },
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<img src=\"attachment:prodom_data_visualization.png\" width=\"250\">"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "colab_type": "text",
    "id": "z0rraiSFNHeZ"
   },
   "source": [
    "<p style=\"background-color:#adebad;\">\n",
    "    2. Demonstrate your implementation of <b> plot_pairwise() </b> below.\n",
    "</p>"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 320
    },
    "colab_type": "code",
    "id": "ELHfd1QrNHea",
    "outputId": "3807f2be-b03e-4c28-e1c1-e9a6725bea7d"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "data.min = 0.0\n",
      "data.median = 11.23\n",
      "data.max = 13.64\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "plot_pairwise(protein[\"dissimilarity\"], protein[\"labels\"])"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<p style=\"background-color:#adebad;\">\n",
    "Read reference [1] and answer the following questions:\n",
    "</p>\n",
    "<p style=\"background-color:#adebad;\">\n",
    "3. Provide a description of the general EM algorithm with main formulas. <br>\n",
    "</p>"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "The EM algorithm includes two calculations and can be summarized as below:\n",
    "\n",
    "1) Firstly, we determine some parameters like initial and final temperature etc.\n",
    "\n",
    "2) Secondly, we initialize the optimal potentials $\\varepsilon^*_{iv}$ and expected assignment probabilities $M_{iv}\\in (0,1)$ and both have the shape ($N\\times K$) where $N$ is the number of samples and $K$ is the number of clusters.\n",
    "\n",
    "3) Then, while final temperature $T_f$ is lower than current temperature $T$:\n",
    " \n",
    "-  Calculate new assignment probabilities $M_{iv}$ and as a function of current potentials $\\varepsilon^*_{iv}$.\n",
    "\n",
    "- Calculate new optimal potentials $\\varepsilon^*_{iv}$ and as a function of current assignments $M_{iv}$.\n",
    "\n",
    "- Do the above two steps until they converge, stop for example when Frobenius norm of the difference between new assignment probabilities and the old one is lower than a threshold. \n",
    "\n",
    "- Update the temperature as $T=T/2$, update the current assignments and potentials with the new ones and go back to the first step.\n",
    "- **Formulas**: Formulas for calculating the variables above can be seen below:\n",
    "\n",
    "\n",
    "1) $$ M_{i \\alpha} = \\frac{\\exp(-\\varepsilon^*_{i\\alpha} / T) }{\\sum^{K}_{v=1} \\exp(-\\varepsilon^*_{iv} / T)} $$\n",
    "\n",
    "\n",
    "2)  $$ \\varepsilon_{iv}^{*} = \\frac{1}{p_vN} \\sum_{k=1}^{N} M_{kv} \\bigg( D_{ik} - \\frac{1}{2p_vN} \\sum_{j=1}^N M_{jv}D_{jk} \\bigg) $$\n",
    "\n",
    "where \n",
    "\n",
    "$$ p_v =\\frac{\\sum_{k=1}^{N} M_{kv}}{N}$$\n",
    "\n",
    "\n",
    "3) $$ X_i = K_i^{-1} \\frac{1}{2} \\sum_{v=1}^K M_{iv} \\big( ||y_v||^2 - \\varepsilon^*_{iv} \\big) \\big(y_v - \\sum_{\\mu =1}^K M_{i\\mu} y_{\\mu}\\big)$$\n",
    "\n",
    "$$ y_v = \\frac{\\sum_{k=1}^N M_{kv} x_k}{\\sum_{k=1}^N M_{kv}}$$\n",
    "\n",
    "where \n",
    "\n",
    "$$ K_i = \\bigg(<yy^T>_i - <y>_i<y^T>_i \\bigg) $$"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<p style=\"background-color:#adebad;\">\n",
    "4. Why is it not possible to apply the EM-scheme for central clustering (algorithm I in [1]) to the case of pairwise clustering? <br>\n",
    "</p>"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    " Algoritm III has the same EM-like procedure above with some additions and can be summarized as below:\n",
    "\n",
    "1) Firstly, we determine some parameters like initial and final temperature etc.\n",
    "\n",
    "2) Secondly, we initialize the embeddings $\\hat{\\mathbf{x}}_i$ with the shape ($N\\times D$), centroids $\\hat{\\mathbf{y}}_v$ with the shape ($K\\times D$) and the expected assignment probabilities $M_{iv}\\in (0,1)$ with the shape ($N\\times K$) where $N$ is the number of samples, $K$ is the number of clusters and $D$ is the number of dimensions of the embeddings.\n",
    "\n",
    "3) Then, while final temperature $T_f$ is lower than current temperature $T$:\n",
    " \n",
    "- Calculate new assignment probabilities $M_{iv}$ as a function of $\\hat{\\mathbf{x}}_i$ and $\\hat{\\mathbf{y}}_v$, where $\\hat{\\mathbf{y}}_v$ corresponds to the centroids.\n",
    "\n",
    "- Until centroids converge; first calculate new embeddings $\\hat{\\mathbf{x}}_i$ with the new $M_{iv}$ and current $\\hat{\\mathbf{y}}_v$ and then find the new centroids by updating $\\hat{\\mathbf{y}}_v$ to fulfill centroid condition.\n",
    "\n",
    "- Do the above two steps until they converge, stop for example when Frobenius norm of the difference between new assignment probabilities and the old one is lower than a threshold. \n",
    "\n",
    "- Update the temperature as $T=T/2$, update the current assignments, embeddings and centroids with the new ones and go back to the first step."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<p style=\"background-color:#adebad;\">\n",
    "5. Describe the idea of mean-field approximation (as KL divergency minimization problem) with the main formulas. How does mean field approximaiton help to solve the problem from the previous question?<br>\n",
    "</p>"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "#Provide your answer here"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<p style=\"background-color:#adebad;\">\n",
    "6. Provide a description of the EM algorithm for pairwise clustering for the case if we have proximity data (algorithm II in [1]). <br>\n",
    "</p>"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "The following approach can be summarized as below:\n",
    "\n",
    "1) Firstly, we determine some parameters like initial and final temperature etc.\n",
    "\n",
    "2) Secondly, we initialize the optimal potentials $\\varepsilon^*_{iv}$ with the shape ($N\\times K$), the embeddings $\\hat{\\mathbf{x}}_i$ with the shape ($N\\times D$), centroids $\\hat{\\mathbf{y}}_v$ with the shape ($K\\times D$) and the expected assignment probabilities $M_{iv}\\in (0,1)$ with the shape ($N\\times K$) where $N$ is the number of samples, $K$ is the number of clusters and $D$ is the number of dimensions of the embeddings.\n",
    "\n",
    "3) Then, while final temperature $T_f$ is lower than current temperature $T$:\n",
    " \n",
    "- Calculate new assignment probabilities $M_{iv}$ and as a function of current potentials $\\varepsilon^*_{iv}$.\n",
    "\n",
    "- Calculate new optimal potentials $\\varepsilon^*_{iv}$ and as a function of current assignments $M_{iv}$.\n",
    "\n",
    "- Do the above two steps until they converge, stop for example when Frobenius norm of the difference between new assignment probabilities and the old one is lower than a threshold. Update the current assignments and potentials with the new ones.\n",
    "\n",
    "- Until centroids converge; first calculate the new centroids $\\hat{\\mathbf{y}}_v$ with $M_{iv}$ and $\\hat{\\mathbf{x}}_i$ and then calculate the new embeddings using new $\\hat{\\mathbf{y}}_v$.\n",
    "\n",
    "- Update the temperature as $T=T/2$ and go back to the first step."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "colab_type": "text",
    "id": "KqWfmZkTNHeg"
   },
   "source": [
    "<p style=\"background-color:#adebad;\">   \n",
    "7. Implement the <b>fit_predict()</b> method without embedding (Algorithm II in [1]) for the template <b>class PairwiseClustering</b>, according to its docstring. You can add more class methods if necessary <br>\n",
    "</p>"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {
    "colab": {},
    "colab_type": "code",
    "id": "zGZWm0t_NHeh"
   },
   "outputs": [],
   "source": [
    "class PairwiseClustering(skl.base.BaseEstimator, skl.base.TransformerMixin):\n",
    "    \"\"\"Template class for PairwiseClustering (PC)\n",
    "    \n",
    "    Attributes:\n",
    "        M (np.ndarray): Assignment probability matrix (n_samples, n_clusters)\n",
    "        centroids (np.ndarray): Embedding centroids (n_clusters, n_components), used only if embed is True.\n",
    "        embeddings (np.ndarray): Embeddings of input samples (n_samples, n_components),\n",
    "                                 used only if embed is True.\n",
    "    Parameters:\n",
    "        n_clusters (int): Maximum number of clusters.\n",
    "        random_state (int): Random seed.\n",
    "        embed (bool): Indicator, if data should be clustered and embedded simultaneously (algorithm III),\n",
    "                      or not (algorithm II).\n",
    "        n_components (int): Dimension of embedding space, used only if embed is True\n",
    "    \"\"\"\n",
    "    \n",
    "    def __init__(self, n_clusters=4, random_state=1, embed=False, n_components=2, T=1, T_f=0.0001, eps=1e-4, rate=0.5, eps_embed=0.05):\n",
    "        self.n_clusters = n_clusters\n",
    "        self.random_state = random_state\n",
    "        self.embed = embed\n",
    "        self.n_components = n_components\n",
    "        self.T = T\n",
    "        self.T_f = T_f\n",
    "        self.eps = eps\n",
    "        self.rate = rate\n",
    "        self.M = None\n",
    "        self.eps_embed = eps_embed\n",
    "        # Add more parameters, if necessary.\n",
    "    \n",
    "    def fit_predict(self, X):\n",
    "        \"\"\"Compute PC for dissimilarity matrix X\n",
    "        \n",
    "        Compute cluster assignments. If embed is True, also compute embeddings/centroids.\n",
    "         \n",
    "        Args:\n",
    "            X (np.ndarray): Dissimilarity matrix (n_samples, n_samples)\n",
    "        \n",
    "        Returns:\n",
    "            C (np.ndarray): Assignments (n_samples,) \n",
    "        \"\"\"\n",
    "        np.random.seed(self.random_state)\n",
    "        \n",
    "        if self.embed:\n",
    "            n_samples    = X.shape[0]\n",
    "            n_clusters   = self.n_clusters\n",
    "            n_components = self.n_components\n",
    "            \n",
    "            # Useful parameters like initial-final temperature and convergence point\n",
    "            T    = self.T     #Initial temperature\n",
    "            T_f  = self.T_f   #Final temperature\n",
    "            eps  = self.eps   #Convergence value\n",
    "            rate = self.rate  #Temperature decrease rate\n",
    "            \n",
    "            # Initialization of the potentials, assignments, embeddings etc.\n",
    "            D   = X.copy()\n",
    "            E   = np.random.uniform(0, 1, size=(n_samples, n_clusters))\n",
    "            M   = np.random.uniform(0, 1, size=(n_samples, n_clusters))\n",
    "            X   = np.random.uniform(0, 1, size=(n_samples, n_components))\n",
    "            y   = np.zeros((n_clusters, n_components))\n",
    "            \n",
    "            while T>T_f:\n",
    "                # EM-like algorithm \n",
    "                while True:\n",
    "                    M_old = M.copy()\n",
    "                    ## E-like step:\n",
    "                    # Exponential trick to prevent numerical errors        \n",
    "                    norms    = -E / T\n",
    "                    max_norm = np.amax(norms, axis=1)[:, np.newaxis]\n",
    "\n",
    "                    # Calculation of optimal assignments\n",
    "                    M = np.exp((-E / T)-max_norm)\n",
    "                    M = M / (np.sum(M, axis=1)[:, np.newaxis])\n",
    "                    \n",
    "                    ## M-like step:\n",
    "                    # Calculation of expected assignment variables\n",
    "                    p_v = (np.sum(M, axis=0)/n_samples)[:, np.newaxis]\n",
    "                    \n",
    "                    # Calculation of optimal potentials for each cluster\n",
    "                    for cluster in range(n_clusters):\n",
    "                        M_c   = ((M[:, cluster])[:, np.newaxis]).T\n",
    "                        temp1 = (1 / (2*p_v[cluster]*n_samples)) * (M_c @ D)\n",
    "                        temp2 = M_c * (D - temp1) * (1 / (p_v[cluster]*n_samples))\n",
    "                        E[:, cluster] = np.sum(temp2, axis=1)\n",
    "\n",
    "                    ## Break the loop if optimal assignments and potentials don't change for the current temperature\n",
    "                    if(np.linalg.norm(M-M_old) < eps): \n",
    "                        break\n",
    "                        \n",
    "                # EM-like algorithm for centroids and embeddings\n",
    "                while True:\n",
    "                    y_old = y.copy()\n",
    "                    \n",
    "                    # Calculation of centroids\n",
    "                    for c in range(n_clusters):\n",
    "                        y[c, :] = np.sum((M[:, c])[:, np.newaxis] * X, axis=0) / np.sum(M[:, c])\n",
    "                    \n",
    "                    # Calculation of embeddings\n",
    "                    for i in range(n_samples):\n",
    "                        # Calculate K_i\n",
    "                        K       = np.cov(y, rowvar=False, ddof=0, aweights=M[i,:])\n",
    "                        # Calculate the terms of K_i*x_i\n",
    "                        tmp1    = y - (M[i, :] @ y)[np.newaxis, :]\n",
    "                        tmp2    = np.diag(y@y.T)[:, np.newaxis] - E[i, :][:, np.newaxis]\n",
    "                        tmp3    = M[i, :][:, np.newaxis] * 0.5\n",
    "                        # Calculate K_i*x_i\n",
    "                        K_X     = np.sum(tmp3 * (tmp2 * tmp1), axis=0)\n",
    "                        # Calculate x_i\n",
    "                        X[i, :] = np.linalg.inv(K) @ K_X\n",
    "                    \n",
    "                    # Normalize x_i\n",
    "                    X = X / n_samples\n",
    "                    \n",
    "                    # Break the loop if centroids converge\n",
    "                    if(np.linalg.norm(y-y_old) < self.eps_embed): \n",
    "                        for c in range(n_clusters):\n",
    "                            y[c, :] = np.sum((M[:, c])[:, np.newaxis] * X, axis=0) / np.sum(M[:, c])\n",
    "                        break\n",
    "                        \n",
    "                ## Update the temperature \n",
    "                T = T*rate\n",
    "            \n",
    "            # Store the necessary attributes\n",
    "            self.X = X\n",
    "            self.y = y\n",
    "            self.M = M\n",
    "            C = np.argmax(M, axis=1)\n",
    "                          \n",
    "            return C\n",
    "        \n",
    "        else:\n",
    "            n_samples  = X.shape[0]\n",
    "            n_clusters = self.n_clusters\n",
    "            \n",
    "            # Useful parameters like initial-final temperature and convergence point\n",
    "            T    = self.T     #Initial temperature\n",
    "            T_f  = self.T_f   #Final temperature\n",
    "            eps  = self.eps   #Convergence value\n",
    "            rate = self.rate  #Temperature decrease rate\n",
    "            \n",
    "            # Initialization of the potentials and assignments\n",
    "            E   = np.random.uniform(0, 1, size=(n_samples, n_clusters))\n",
    "            M   = np.random.uniform(0, 1, size=(n_samples, n_clusters))\n",
    "            \n",
    "            \n",
    "            # EM-like algorithm\n",
    "            while T>T_f:\n",
    "                while True:\n",
    "                    M_old = M.copy()\n",
    "                    ## E-like step:\n",
    "                    # Exponential trick to prevent numerical errors        \n",
    "                    norms    = -E / T\n",
    "                    max_norm = np.amax(norms, axis=1)[:, np.newaxis]\n",
    "\n",
    "                    # Calculation of optimal assignments\n",
    "                    M = np.exp((-E / T)-max_norm)\n",
    "                    M = M / (np.sum(M, axis=1)[:, np.newaxis])\n",
    "                    \n",
    "                    ## M-like step:\n",
    "                    # Calculation of expected assignment variables\n",
    "                    p_v = (np.sum(M, axis=0)/n_samples)[:, np.newaxis]\n",
    "                    \n",
    "                    # Calculation of optimal potentials for each cluster\n",
    "                    for cluster in range(n_clusters):\n",
    "                        M_c   = ((M[:, cluster])[:, np.newaxis]).T\n",
    "                        temp1 = (1 / (2*p_v[cluster]*n_samples)) * (M_c @ X)\n",
    "                        temp2 = M_c * (X - temp1) * (1 / (p_v[cluster]*n_samples))\n",
    "                        E[:, cluster] = np.sum(temp2, axis=1)\n",
    "\n",
    "                    ## Break the loop if optimal assignments and potentials don't change for the current temperature\n",
    "                    if(np.linalg.norm(M-M_old) < eps): \n",
    "                        break\n",
    "                        \n",
    "                ## Update the temperature        \n",
    "                T = T*rate\n",
    "            \n",
    "            # Store the necessary attributes\n",
    "            self.M = M\n",
    "            C = np.argmax(M, axis=1)\n",
    "                  \n",
    "            return C\n",
    "    \n",
    "    \n",
    "    def cost(self, D, M):\n",
    "        \"\"\"Compute the costs of the clustering solution M for dissimilarity D.\n",
    "        \n",
    "        The costs refer to the pairwise clustering costs in eq. (16) of reference [1].\n",
    "        \n",
    "        Args:\n",
    "            D (np.ndarray): Dissimilarity matrix (n_samples, n_samples)\n",
    "            M (np.ndarray): Assignment probability matrix (n_samples, n_clusters)\n",
    "            \n",
    "        Returns:\n",
    "            cost (float): Costs of current clustering\n",
    "        \"\"\"\n",
    "        \n",
    "        # Necessary variables\n",
    "        n_samples = D.shape[0]\n",
    "        C         = np.argmax(M, axis=1)      \n",
    "        p_v       = np.zeros((self.n_clusters,))\n",
    "        \n",
    "        ## Cost calculation:\n",
    "        for c in range(self.n_clusters):\n",
    "            p_v[c] = np.sum(C==c)/n_samples\n",
    "        \n",
    "        s = np.zeros((n_samples, n_samples))\n",
    "        for c in range(self.n_clusters):\n",
    "            tmp = ((C==c)*1)[:, np.newaxis] @ ((C==c)*1)[np.newaxis, :]\n",
    "            s   = s + tmp*(1/p_v[c])\n",
    "                \n",
    "        cost = np.sum((s - 1) * D / (2*n_samples))\n",
    "                \n",
    "        return cost"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "colab_type": "text",
    "id": "pwJA1TSqNHek"
   },
   "source": [
    "<p style=\"background-color:#adebad;\">\n",
    "8. Create an instance of your Pairwise Clustering (PC) class, then fit and predict the Protein data. <br>\n",
    "9. Record the execution time.\n",
    "</p>"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 51
    },
    "colab_type": "code",
    "id": "dJhRnnHfNHel",
    "outputId": "dc1bae15-3a7c-4e61-ede5-7a309b9b150d"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "CPU times: user 1.07 s, sys: 8.06 ms, total: 1.08 s\n",
      "Wall time: 312 ms\n"
     ]
    }
   ],
   "source": [
    "%%time\n",
    "PC = PairwiseClustering()\n",
    "labels_pair = PC.fit_predict(protein[\"dissimilarity\"])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "CPU times: user 224 ms, sys: 77.3 ms, total: 302 ms\n",
      "Wall time: 551 ms\n"
     ]
    }
   ],
   "source": [
    "%%time\n",
    "SC = skl.cluster.SpectralClustering(n_clusters=4, random_state=42, affinity=\"precomputed\")\n",
    "affinities = np.exp(- protein[\"dissimilarity\"] / 3) # SC takes positive affinities as input\n",
    "labels_spectral= SC.fit_predict(affinities)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "colab_type": "text",
    "id": "SNn7EbPzNHer"
   },
   "source": [
    "<h2 style=\"background-color:#f0b375;\">\n",
    "Section 4.5 \n",
    "<span style=font-size:50%> Complete all problems in this (4.5.1-4.5.5) and previous sections to get a grade of 4.5 </span>\n",
    "</h2>"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "In this section, you have to compare pairwise clustering with another clustering approach, that also works with proximity data, in terms of different metrics."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<p style=\"background-color:#adebad;\">\n",
    "1. Read https://scikit-learn.org/stable/modules/clustering.html#spectral-clustering and shortly recap the idea of spectral clustering.\n",
    "</p>"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {},
   "outputs": [],
   "source": [
    "#Provide your answer here\n",
    "from sklearn.cluster import SpectralClustering"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<p style=\"background-color:#adebad;\">\n",
    "2. Predict the Protein data with sklearn.SpectralClustering. <br>\n",
    "3. Record the execution time. <br>\n",
    "</p>"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "CPU times: user 58.3 ms, sys: 3.63 ms, total: 61.9 ms\n",
      "Wall time: 56.1 ms\n"
     ]
    }
   ],
   "source": [
    "%%time\n",
    "SC = skl.cluster.SpectralClustering(n_clusters=4, random_state=42, affinity=\"precomputed\")\n",
    "affinities = np.exp(- protein[\"dissimilarity\"] / 3) # SC takes positive affinities as input\n",
    "labels_spectral= SC.fit_predict(affinities)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<p style=\"background-color:#adebad;\">\n",
    "4. Complete the function <b>compare_clusterings()</b> and use it to compare the spectral and the pairwise clustering methods. <br>\n",
    "Use the following metrics to compare clusterings:\n",
    "</p>\n",
    "<ul style=\"background-color:#adebad;\"> \n",
    "<li> V-measure cluster labeling given a ground truth. </li>\n",
    "<li> Normalized Mutual Information between two clusterings. </li>\n",
    "<li> Homogeneity metric of a cluster labeling given a ground truth. </li>\n",
    "<li> The Fowlkes-Mallows score </li>\n",
    "<li> Rand index adjusted for chance. </li>\n",
    "<li> Adjusted Mutual Information between two clusterings. </li>\n",
    "</ul>\n",
    "<p style=\"background-color:#adebad;\">\n",
    "Hint: use <b>sklearn.metrics.v_measure_score(),sklearn.metrics.normalized_mutual_info_score(), sklearn.metrics.homogeneity_score(), sklearn.metrics.fowlkes_mallows_score(), sklearn.metrics.completeness_score(), sklearn.metrics.adjusted_rand_score(), sklearn.metrics.adjusted_mutual_info_score()</b>. You can also find explanations and examples there: https://scikit-learn.org/stable/modules/classes.html?highlight=metrics#module-sklearn.metrics\n",
    "</p>"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {},
   "outputs": [],
   "source": [
    "## This part is just for creating proper legend labels\n",
    "def toLabel(key):\n",
    "    if(len(key) > 19):\n",
    "        res = key[20:]\n",
    "        if res == \"1\":\n",
    "            return \"a=1\"\n",
    "        elif res == \"2\":\n",
    "            return \"a=2\"\n",
    "        elif res == \"1_2\":\n",
    "            return \"a=1/2\"\n",
    "        elif res == \"log\":\n",
    "            return \"a=log\"\n",
    "        elif res ==  \"exp\":\n",
    "            return \"a=exp\"\n",
    "        elif res == \"Transform1\":\n",
    "            return \"Best of Transform 1\"\n",
    "        elif res == \"Transform2\":\n",
    "            return \"Best of Transform 2\"\n",
    "        elif res == \"Transform3\":\n",
    "            return \"Best of Custom Transform\"\n",
    "    else:\n",
    "        return key"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {
    "colab": {},
    "colab_type": "code",
    "id": "Nyr9fUgfNHev"
   },
   "outputs": [],
   "source": [
    "def compare_clusterings(labels_true, width=0.35, **labels_pred):\n",
    "    \"\"\"Show a summary of different clustering metrics\n",
    "    \n",
    "    Args:\n",
    "        labels_true (array): Ground-truth cluster labels\n",
    "    \n",
    "    Note:\n",
    "        Each kwarg should be an array with predicted labels\n",
    "    \"\"\"\n",
    "    n_pred = len(labels_pred)\n",
    "    \n",
    "    scores = []\n",
    "    for name in dir(skl.metrics.cluster):\n",
    "        score = getattr(skl.metrics.cluster, name)\n",
    "        try:\n",
    "            score(labels_true=[0], labels_pred=[0]) # Check if score takes the right arguments\n",
    "            if \"score\" not in score.__name__:\n",
    "                raise TypeError\n",
    "            if score.__name__ == \"mutual_info_score\": # Exclude, not normalized to 1.0\n",
    "                raise TypeError\n",
    "        except TypeError:\n",
    "            continue\n",
    "        else:\n",
    "            scores.append(score)\n",
    "    \n",
    "    ## Calculate the scores for each clustering methods\n",
    "    clustering_scores = {}\n",
    "    for key, val in labels_pred.items():\n",
    "        # First match labels with the ground truth labels\n",
    "        n_clusters = len(np.unique(labels_true))\n",
    "        labels     = np.zeros_like(val)\n",
    "\n",
    "        for i in range(n_clusters):\n",
    "            mask         = val==i\n",
    "            labels[mask] = mode(labels_true[mask])[0]\n",
    "\n",
    "        permuted_preds   = labels\n",
    "        \n",
    "        # Calculate the score for each function\n",
    "        func_scores = {}\n",
    "        for func in scores:\n",
    "            func_scores[func.__name__] = func(labels_true=labels_true, labels_pred=permuted_preds)\n",
    "        \n",
    "        clustering_scores[key] = func_scores\n",
    "    \n",
    "    func_names = ['v_measure_score', 'normalized_mutual_info_score', 'homogeneity_score', 'fowlkes_mallows_score',\n",
    "                  'completeness_score', 'adjusted_rand_score', 'adjusted_mutual_info_score']\n",
    "    \n",
    "    ## Show the scores in a DataFrame table\n",
    "    data = []\n",
    "    index = []\n",
    "    for key, val in clustering_scores.items():\n",
    "        data.append(val)\n",
    "        index.append(toLabel(key))\n",
    "    df = pd.DataFrame(data, index=index)\n",
    "    df = df[func_names]\n",
    "    \n",
    "    # Calculate and print how many times each param is the best for different function scores\n",
    "    n_tr = len(df.index)\n",
    "    cnts = {}\n",
    "    for i in range(n_tr):\n",
    "        t1 = df.iloc[i%n_tr].values >= df.iloc[(i+1)%n_tr].values\n",
    "        t2 = df.iloc[i%n_tr].values >= df.iloc[(i+2)%n_tr].values\n",
    "        c  = np.logical_and(t1, t2)\n",
    "        print(df.index[i] + \" has the highest score for \" + str(np.sum(c)) + \" different functions.\")\n",
    "        cnts[df.index[i]] = np.sum(c)\n",
    "        \n",
    "    # Print the best parameter according to the ranking\n",
    "    print(\"The winner is \" + max(cnts, key=cnts.get) + \".\")\n",
    "    \n",
    "    # Print the data frame\n",
    "    display(df.T) \n",
    "    \n",
    "    # Plot the scores\n",
    "    N = len(func_names)\n",
    "    indices = np.flip(np.arange(N))\n",
    "    j = 0\n",
    "    colors = [\"darkorange\", \"deepskyblue\", \"firebrick\"]\n",
    "    for key, val in clustering_scores.items():\n",
    "        i = 0\n",
    "        for func_key, ind in zip(func_names, indices):\n",
    "            plt.barh(ind+width*j, val[func_key], width, color=colors[j], label=toLabel(key) if i == 0 else \"\")\n",
    "            i = i+1\n",
    "        j = j+1\n",
    "\n",
    "    plt.yticks(indices - width/2 + (j)*width / (2), tuple(func_names))\n",
    "    plt.xlim([0, 1.03])\n",
    "    plt.xticks(np.arange(0, 1.01, 0.2))\n",
    "    plt.legend(bbox_to_anchor=(1, 1), loc='upper left', ncol=1)\n",
    "    plt.show()\n",
    "    \n",
    "    return max(cnts, key=cnts.get)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "colab_type": "text",
    "id": "GbeFm900NHez"
   },
   "source": [
    "<p style=\"background-color:#adebad;\">\n",
    "    5. Demonstrate your implementation of <b>compare_clusterings()</b>. See example of how the output should look like below. You should produce one plot. Hint: use <b>matplotlib.pyplot.barh(), matplotlib.pyplot.xticks(), matplotlib.pyplot.yticks()</b> to get horizontal histogram for different metrics as shown below.\n",
    "\n",
    "</p>"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 552
    },
    "colab_type": "code",
    "id": "f7ADfUYhNHe4",
    "outputId": "aad3c6c7-4e0f-48b3-ed96-61bdb142f01d"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "SpectralClustering has the highest score for 0 different functions.\n",
      "GroundTruth has the highest score for 7 different functions.\n",
      "The winner is GroundTruth.\n"
     ]
    },
    {
     "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>SpectralClustering</th>\n",
       "      <th>GroundTruth</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>v_measure_score</th>\n",
       "      <td>0.787844</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>normalized_mutual_info_score</th>\n",
       "      <td>0.787844</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>homogeneity_score</th>\n",
       "      <td>0.735734</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>fowlkes_mallows_score</th>\n",
       "      <td>0.830741</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>completeness_score</th>\n",
       "      <td>0.847899</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>adjusted_rand_score</th>\n",
       "      <td>0.746487</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>adjusted_mutual_info_score</th>\n",
       "      <td>0.783860</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                              SpectralClustering  GroundTruth\n",
       "v_measure_score                         0.787844          1.0\n",
       "normalized_mutual_info_score            0.787844          1.0\n",
       "homogeneity_score                       0.735734          1.0\n",
       "fowlkes_mallows_score                   0.830741          1.0\n",
       "completeness_score                      0.847899          1.0\n",
       "adjusted_rand_score                     0.746487          1.0\n",
       "adjusted_mutual_info_score              0.783860          1.0"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "#Example how output sould looks like after the following command\n",
    "_ = compare_clusterings(labels_true=protein[\"labels\"],\n",
    "                        SpectralClustering=labels_spectral,\n",
    "                        GroundTruth=protein[\"labels\"])"
   ]
  },
  {
   "attachments": {
    "data_visualiztion.png": {
     "image/png": "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"
    }
   },
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<img src=\"attachment:data_visualiztion.png\" width=\"700\">"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "colab_type": "text",
    "id": "eMjveOeJNHfA"
   },
   "source": [
    "<h2 style=\"background-color:#f0b375;\">\n",
    "Section 5.0 \n",
    "<span style=font-size:50%> Complete all problems in this (5.0.1-5.0.5) and previous sections to get a grade of 5.0 </span>\n",
    "</h2>"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "In this section, you have to investigate the quality of clustering depending on various transformations from similarity data to dissimilarity data and compare results, in terms of different metrics."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<p style=\"background-color:#adebad;\">\n",
    "    As PairwiseClustering takes dissimilarities as input, we need to transform the similarity data from <b>Prodom</b> accordingly.<br>\n",
    "    1. Run <b>fit_predict()</b> with all the transformations and parameters below. Run compare_clusterings for each set of parameter values for your <b>PairwiseClustering</b> algorithm. <br>\n",
    "    2. Produce the output of <b>compare_clusterings()</b> for transformation families below. You should get 3 plots: 1st transformation family comparison, 2nd transformation family comparison, then comparison between of the best transformations of every family.<br>\n",
    "<b>NOTE</b>: The best transformation in a family is the one that has the maximum average of all the scores above.\n",
    "</p>"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "a=1 has the highest score for 7 different functions.\n",
      "a=2 has the highest score for 0 different functions.\n",
      "a=exp has the highest score for 0 different functions.\n",
      "The winner is a=1.\n"
     ]
    },
    {
     "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>a=1</th>\n",
       "      <th>a=2</th>\n",
       "      <th>a=exp</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>v_measure_score</th>\n",
       "      <td>0.210681</td>\n",
       "      <td>0.150436</td>\n",
       "      <td>0.150436</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>normalized_mutual_info_score</th>\n",
       "      <td>0.210681</td>\n",
       "      <td>0.150436</td>\n",
       "      <td>0.150436</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>homogeneity_score</th>\n",
       "      <td>0.144479</td>\n",
       "      <td>0.094544</td>\n",
       "      <td>0.094544</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>fowlkes_mallows_score</th>\n",
       "      <td>0.525048</td>\n",
       "      <td>0.522191</td>\n",
       "      <td>0.522191</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>completeness_score</th>\n",
       "      <td>0.388861</td>\n",
       "      <td>0.367969</td>\n",
       "      <td>0.367969</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>adjusted_rand_score</th>\n",
       "      <td>0.061500</td>\n",
       "      <td>0.026986</td>\n",
       "      <td>0.026986</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>adjusted_mutual_info_score</th>\n",
       "      <td>0.209615</td>\n",
       "      <td>0.149813</td>\n",
       "      <td>0.149813</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                                   a=1       a=2     a=exp\n",
       "v_measure_score               0.210681  0.150436  0.150436\n",
       "normalized_mutual_info_score  0.210681  0.150436  0.150436\n",
       "homogeneity_score             0.144479  0.094544  0.094544\n",
       "fowlkes_mallows_score         0.525048  0.522191  0.522191\n",
       "completeness_score            0.388861  0.367969  0.367969\n",
       "adjusted_rand_score           0.061500  0.026986  0.026986\n",
       "adjusted_mutual_info_score    0.209615  0.149813  0.149813"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "#Transformation family 1\n",
    "# D = ((max(S) - S) / (max(S) - min(S))) ** 1\n",
    "# D = ((max(S) - S) / (max(S) - min(S))) ** 2\n",
    "# D = np.exp(((max(S) - S) / (max(S) - min(S))))\n",
    "\n",
    "#Please add your code here\n",
    "## Calculation of different dissimilarity matrices with different transformations of the similarity matrix\n",
    "S  = prodom[\"similarity\"]\n",
    "D1 = ((np.max(S) - S) / (np.max(S) - np.min(S))) ** 1\n",
    "D2 = ((np.max(S) - S) / (np.max(S) - np.min(S))) ** 2\n",
    "D3 = np.exp((np.max(S) - S) / (np.max(S) - np.min(S)))\n",
    "D  = [D1, D2, D3]\n",
    "\n",
    "# Calculation of different cluster predictions with different dissimilarity matrices\n",
    "labels1_SC = []\n",
    "for i in range(3):\n",
    "    SC = skl.cluster.SpectralClustering(n_clusters=4, random_state=42, affinity=\"precomputed\")\n",
    "    affinities = np.exp(- D[i] / 3) # SC takes positive affinities as input\n",
    "    labels1_SC.append(SC.fit_predict(affinities))\n",
    "    \n",
    "# Print and plot the scores for different functions and different params\n",
    "best_transform1_SC = compare_clusterings(labels_true=prodom[\"labels\"],\n",
    "                                         width = 0.2,\n",
    "                                         SpectralClustering_a1=labels1_SC[0],\n",
    "                                         SpectralClustering_a2=labels1_SC[1],\n",
    "                                         SpectralClustering_aexp=labels1_SC[2])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "a=1 has the highest score for 0 different functions.\n",
      "a=1/2 has the highest score for 7 different functions.\n",
      "a=log has the highest score for 0 different functions.\n",
      "The winner is a=1/2.\n"
     ]
    },
    {
     "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>a=1</th>\n",
       "      <th>a=1/2</th>\n",
       "      <th>a=log</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>v_measure_score</th>\n",
       "      <td>0.182547</td>\n",
       "      <td>0.385255</td>\n",
       "      <td>0.376993</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>normalized_mutual_info_score</th>\n",
       "      <td>0.182547</td>\n",
       "      <td>0.385255</td>\n",
       "      <td>0.376993</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>homogeneity_score</th>\n",
       "      <td>0.121589</td>\n",
       "      <td>0.301001</td>\n",
       "      <td>0.292954</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>fowlkes_mallows_score</th>\n",
       "      <td>0.488250</td>\n",
       "      <td>0.562150</td>\n",
       "      <td>0.559588</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>completeness_score</th>\n",
       "      <td>0.366078</td>\n",
       "      <td>0.535012</td>\n",
       "      <td>0.528642</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>adjusted_rand_score</th>\n",
       "      <td>0.008838</td>\n",
       "      <td>0.202437</td>\n",
       "      <td>0.194235</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>adjusted_mutual_info_score</th>\n",
       "      <td>0.181983</td>\n",
       "      <td>0.384531</td>\n",
       "      <td>0.376254</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                                   a=1     a=1/2     a=log\n",
       "v_measure_score               0.182547  0.385255  0.376993\n",
       "normalized_mutual_info_score  0.182547  0.385255  0.376993\n",
       "homogeneity_score             0.121589  0.301001  0.292954\n",
       "fowlkes_mallows_score         0.488250  0.562150  0.559588\n",
       "completeness_score            0.366078  0.535012  0.528642\n",
       "adjusted_rand_score           0.008838  0.202437  0.194235\n",
       "adjusted_mutual_info_score    0.181983  0.384531  0.376254"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAkMAAAD4CAYAAADvq+IEAAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADh0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uMy4xLjMsIGh0dHA6Ly9tYXRwbG90bGliLm9yZy+AADFEAAAgAElEQVR4nO3de5xWVdn/8c+Xg4IyQQqR5gGdOCgqGCNKplCRWYRpnio1wVOUeCiw/OVjqaVFlqZlKpqgRmWWqYiKRgKGSs7IYcAUBel5LFFUQBFEgev3x16jd8OcmZkb2N/36zWv2fe6117r2uueYS7W2gdFBGZmZmZ51abYAZiZmZkVk5MhMzMzyzUnQ2ZmZpZrTobMzMws15wMmZmZWa61K3YAZjXp2rVr9OjRo9hhmJltVSoqKl6NiG7FjmNr42TItkg9evSgvLy82GGYmW1VJP2r2DFsjbxMZmZmZrnmZMjMzMxyzcmQmZmZ5ZrPGTIzM9uGVVRUfKhdu3Y3A/uRz0mQjcCC9evXnzFgwIBXaqrgZMjMzGwb1q5du5s//OEP79OtW7cVbdq0yd0DSTdu3Kjly5fvu2zZspuBo2qqk8cM0czMLE/269at2xt5TIQA2rRpE926dVtFNjNWc51WjMfMzMxaX5u8JkJV0vHXmvM4GTIzM7Nc8zlDtkVaVVnJ5NLSYodhLWz44sXFDsEsf36uAc3a3pioaNb2kgceeKDTmDFjdl+0aNEON91005KRI0euaIl+wDNDZmZmtgXae++935kwYcLS4cOHv9bSfXlmyMzMzFrU0KFDS1966aXt1q1b12bUqFEvjx079tX69undu/c7AG3atPy8jZMhMzMza1GTJk1a2r179w2rV6/WgQceuO/JJ5+84vTTT99z8eLFHarXHT169MujR49u8dmgQk6GzMzMrEWNGzeu+5QpU7oALFu2rP3ChQs7TJkyZUmx46riZMjMzMxazH333VcyY8aMkvLy8mdKSko2Dhw4sPfatWvbDBs2bG/PDJnV4fke+3PUjeXFDsNa2vT3N2NIsYIws5a0cuXKtp07d95QUlKycc6cOR3mzZu3I4BnhqzVSGoXEevz2r+ZmVXTQpfC1+bYY49dNX78+G69evXat7S09O1+/fq91ZD9ZsyYscMJJ5zw0TfeeKPttGnTulx++eW7Pv/88wtbIkYnQw0gaRzwr4j4dXp9CfBmRPy8Wr0hwKXAy0B/4C6gEjgP6AgcHRGLJXUDbgD2SLueHxGzJA0EfpHqrgVGRsSzkvoCE4DtyG6HcCzwLnBfROyX+h4LdIqISyRNBx4DDgXulXRbTf3VcqyDgWvSywAOj4g3JX0HOIXsgXcPRMSFkvqndncAFgOnRcSKzenfzMy2LR07doyZM2c+19j9Bg8evObll1+e3xIxVedkqGH+QJak/Dq9PgE4spa6/YB9gNeBJcDNETFQ0nnAOcD5ZMnG1RHxd0l7AFPTPs+QJR/rJQ0FriBLfEYB10TEJEnbAW2B7vXE3CUiBgNI+l0t/dVkLHB2Ss46AW9L+hxwNHBwRKyRtFOqextwTkTMkHQZ8IN0fE3qX9JZwFkAdN+j+ttmZmYtwslQA0TEHEkfkrQr0A1YERH/W0v1JyPiJQBJi4GHUnkl8Mm0PRTYV1LVPh+QVAJ0Bm6V1JNsVqZ9ev9x4CJJuwF3RcRzBfvW5o6C7Rr7i4g3a9hvFnCVpEmprxdTYjYhItak8XhdUmeyhGdG2u9W4M7N6T8ixgPjAdS7LNfP0TEzs9bjZKjh/gQcB3yYbKaoNusKtjcWvN7I++PdBhgUEWsLd5T0S+CRiDhGUg/S6aUR8TtJs4FhwFRJZwCL+O87iFc/I79wTbbG/moSET+RNAX4PPBESoRElpw1RpP6NzMza21OhhruD8BNQFdg8Ga29RAwGrgSQFL/iJhLNjP071RnRFVlSXsDSyLi2rR9APAo8CFJOwOrgS8ADzayv01IKo2ISqBS0iCgT9r/+5J+V7VMlmaHVkg6LCIeJTufaEZNbTam/yoD1lRQXlHv7Ne2YYwnwczMisnPJmugiFgIlAD/rloG2wznAmWS5kt6muycIICfAj+WNIvsvKAqJwILJM0lS05ui4h3gcuA2cB9ZOcbNba/mpwvaYGkeWQncT8QEQ8C9wLlKYaxqe6pwJWS5pOdMH5ZM/RvZmbWqhTh/5Xalqdsd0X5+fXX2yZ4ZsjMmomkiogoKyybN2/e0n79+tX7LLBt3bx587r269evR03veZnMzMwsRzSdAc3ZXgyhRe5b9MADD3QaM2bM7osWLdrhpptuWjJy5MgVhe8fdthhPW+77bal55133m7z58/fsX379tG/f/+3fvvb3/5r++23b9T/Mr1M1gSS9pc0t9rX7GLH1RiSRtZwDNcVOy4zMzOAvffe+50JEyYsHT58+CaP5li9erVWrlzZbq+99nr3pJNOen3JkiULnn322YVvv/22fvGLX3RtbF+eGWqCdIJx/2LHsTkiYgLZjRzNzMxa1NChQ0tfeuml7datW9dm1KhRL48dO7beZbvevXu/A9CmzabzNvfff3/JoYce+ibAiSeeuKqqvKys7K0XX3xxu8bG52TItkzdB8AYP5vMzGxbMGnSpKXdu3ffsHr1ah144IH7nnzyyStOP/30PZv6oNb777+/85e+9KWVhWXr1q3THXfcsfNVV131f42Nz8mQmZmZtahx48Z1nzJlSheAZcuWtV+4cGGHzXlQ65NPPtnphhtueLGw7NRTT93jkEMOWX3kkUeubmx7TobMzMysxdx3330lM2bMKCkvL3+mpKRk48CBA3uvXbu2zbBhw/ZuyszQ008/vd0uu+zyTocOHd47SXrMmDG7vPrqq+2mTp26uCkxOhkyMzOzFrNy5cq2nTt33lBSUrJxzpw5HebNm7cjQFNnhu65557ORxxxxBtVr6+66qquf/vb3zo/+uijz7Zt27auXWvlZMjMzCxHWupS+Noce+yxq8aPH9+tV69e+5aWlr7dr1+/t+rfC2bMmLHDCSec8NE33nij7bRp07pcfvnluz7//PMLH3744c7XX3/9e88H/c53vrPnLrvssq6srGwfgC984QsrfvaznzXq5shOhszMzKzFdOzYMWbOnPlcY/cbPHjwmpdffnl+YdnatWv1yiuvtK+60gxg/fr1m53c+T5DZmZmtlXo2LFjLFiw4J/N3a6TITMzM8s1J0NmZmaWa06GzMzMLNecDJmZmVmuORkyMzOzXPOl9bZFWlVZyeTS0mKHYS1s+OIm3SzWzDbD5NLSAc3Z3vDFi1vkvkXf/va3d+3UqdOGyy677OWWaL+QZ4bMzMws15wMmZmZWYsaOnRoad++fff56Ec/2vdnP/tZ18bu/9hjj3Xs169fn169eu37mc98pnT58uVtIbtLda9evfbt379/n69//eu79ezZs29T4nMyZGZmZi1q0qRJSxcuXPjPuXPnPn3jjTd2X7ZsWdthw4bt3adPn32rf/3qV7/aufr+I0aM2OuKK654cdGiRU/37dt37Xe/+91dAc4444y9rrvuun/NnTv3mbZt28amPTeMzxkyMzOzFjVu3LjuU6ZM6QKwbNmy9gsXLuzQ0Ae1vvbaa23ffPPNtsOGDVsNcOaZZ752/PHH7/3qq6+2feutt9p85jOfeQvg1FNPff3hhx/u0pT4nAyZmZlZi7nvvvtKZsyYUVJeXv5MSUnJxoEDB/Zeu3Ztm2HDhu29ePHiDtXrjx49+uXRo0e/Vl+7EU2eCNqEkyHbIj3fY3+OurG82GFYS5v+/mYMKVYQZtaSVq5c2bZz584bSkpKNs6ZM6fDvHnzdgRo6MzQzjvvvOEDH/jAhgcffLDTkUceufo3v/nNzoMGDVrdrVu3DTvuuOPGadOm7fjpT3/6rdtvv32npsa4xSdDkqYDYyOiXNL9wFcjYuVmtDcktfeFZgqxpvbfiYjHNqON1RHRqZb3dgWujYjj6mnjeOAyYFlEfLKpsZiZ2balpS6Fr82xxx67avz48d169eq1b2lp6dv9+vV7q7FtTJgw4YVvfOMbe5577rlt9thjj3W///3vlwLceOONS0eNGrXnDjvssPHQQw99s6SkZENTYmzRZEhSu4hY31ztRcTnm6utFjQEWA00ORmqS0T8B6gzEUpOB74ZEY+0RBwN1dw/A2ZmtnXp2LFjzJw587nG7nfVVVf9p2r74x//+Np58+Y9U73OgAED1i5atOhpgO9973sfbkqiBQ24mkxSD0n/lHSTpIWSHpLUUVJ/SU9Imi/pL5I+mOpPl3SFpBnAeZImSrpe0iOSlkgaLOmW1ObEgn6ul1Se+ri0lliWSuoqaZSkuenrBUmPpPePkPS4pKck3SmpUyo/UtIzkv4OfKme471E0q3pOJdK+pKkn0qqlPSgpPaFsaTtsnTcPYBRwLdSbIel4z+uoP3V6XsnSdNSrJWSvljfZ1HweSxI2yMk3ZXiek7ST1P594FPADdIulJSB0kTUj9zJNU6UySpr6R/pPjnS+qZyr+WXs+TdHsq2zMdw/z0fY9UPlHSVelzGSdpx/SZP5n6r/FYJZ2VfgbKWbW8IcNhZmY59sc//rFznz599u3Zs2ffxx57rNPll1/+UlPaaeil9T2B6yKiL7ASOBa4DfhuRBwAVAI/KKjfJSIGR8TP0+sPAp8CvgVMBq4G+gL7S+qf6lwUEWXAAcBgSQfUFkxE3BAR/YGDgBeBq1Ji8j/A0Ij4GFAOfFtSB+AmYDhwGPDhBhxvKTAM+CLwW+CRiNgfWJvKa4trKXADcHVE9I+IR+vo423gmBTrJ4GfS1IDYquuP3AisD9woqTdI+IysuM/KSIuAM5O8e0PfAW4NY1LTUYB16TxLQNelNQXuAj4VET0A85LdX8F3JZ+BiYB1xa004vssxiT9v1bRByUjvVKSTtW7zgixkdEWUSU0blbE4bCzMzy5Mwzz1zxzDPPPP3cc88tnD59+vO77rprk1YiGpoMvRARc9N2BVmy0CUiZqSyW4HDC+rfUW3/yZGd9l0JvBwRlRGxEVgI9Eh1TpD0FDCHLFHatwFxXUP2R3YycEjaZ5akucCpwJ5AnxT/cymG3zag3Qci4t0Ub1vgwVReWRDv5hJwhaT5wF+BjwDdm9DOtIhYFRFvA0+THXN1nwBuB4iIZ4B/kSUrNXkc+J6k7wJ7RsRaskT2TxHxamrj9VR3EPC7tH176qfKnRFRtXZ7BHBh+lymAx2APRp7oGZm1iQbN27c2JT/bG8z0vFvrO39hp4ztK5gewNQ33X81dfsqvbfWK2tjUA7SXsBY4GDImJFWj6rbeYCyJaIyP7wj64qAh6OiK9Uq9cfaOz1d+sAImKjpHfj/ev3NvL+mK3n/WSyrljfq5dmfrZL5ScB3YABEfGupKX1tFNnrMkGav5MG/xLEBG/kzSbbAZsqqQz0v4NGcPCOoU/AwKOjYhnGxrHgDUVlFfk5Hd3TPNdHmpmVoMFy5cv37dbt26r2rRpk7t/cDZu3Kjly5d3BhbUVqepJ1CvAlZIOiwtBZ0CzKhnn7p8gOyP5ypJ3YHP8V8X3f43SQPIkqfD0gwTwBPAdZI+GhHPS9oB2A14BthLUmlELCZbJmoOS4EBwANky4ZV3kzHU73eH8mW3dqn8s7AKykR+iQ1z+g0l5lkydffJPUim5WpMTGRtDewJCKuTdsHAA8Df5F0dUS8JmmnNDv0GPBlslmhk4C/19L/VOAcSedEREg6MCLmNOsRmplZjdavX3/GsmXLbl62bNl+5PPJExuBBevXrz+jtgqbczXZqWQn6O4ALAFGNrWhiJgnaQ7ZstkSYFY9u4wGdgIeSafZlEfEGWm26PeStk/1/iciFkk6C5gi6VWyP9j7NTXWApcCv5H0PWB2Qflk4E/pJOFzyM5XukfSP4BpvD9jMgmYLKkcmEuWtLWUX5N9VpVkM1UjImJdLXVPBE6W9C6wDLgsIl6XdDkwQ9IGsqXMEcC5wC2SLgCWU/vPwA+BXwDz0+zYUqBFbm1gZmb/bcCAAa8ARxU7ji2ZmvMOjmbNpWx3Rfn5xY6ilXiZzMyaiaSKdDGSNUIep8vMzMzM3rPF34G6pUgayfuXiFeZFRFnFyOe6iTtT7oCrMC6iDi4mdr/LDCuWvELEXFMc7RvZma2tfAymW2RysrKorzczyYzM2sML5M1jZfJzMzMLNecDJmZmVmuORkyMzOzXHMyZGZmZrnmZMjMzMxyzcmQmZmZ5ZqTITMzM8s1J0NmZmaWa06GzMzMLNecDJmZmVmuORkyMzOzXMvtg1pty7aqspLJpaXFDiOXhi9eXOwQzMxalWeGzMzMLNecDJmZmVmuORkyMzOzXHMyZGZmZrnmZMjMzMxyTRFR7BjMNqHeZcGN5cUOI/diSLEjMLPGkFQREWXFjmNrk9uZIUk9JC0odhzNSdJj6XsPSV8tdjxmZmZbg9wmQ9uiiPh42uwBbBHJkCTfy8rMzLZoeU+G2kq6SdJCSQ9J6iipv6QnJM2X9BdJHwSQNF3S1ZJmSvqnpIMk3SXpOUk/qmpQ0rclLUhf5xeUXyzpGUkPS/q9pLGpvFTSg5IqJD0qqU8qnyjpWkmPSVoi6biCti6Q9GSK8dKC8tVp8yfAYZLmSvpWard/Qb1Zkg6oaUAkDU77zZU0R1JJKv+OpEpJ8yT9JJXVNVZXSJoBnCepm6Q/p5iflHRoLX2fJalcUjmrljfqgzQzM2uqvCdDPYHrIqIvsBI4FrgN+G5EHABUAj8oqP9ORBwO3ADcA5wN7AeMkLSzpAHASOBg4BDgTEkHSipLbR8IfAkoXM8dD5wTEQOAscCvC97bBfgE8AWyBAdJR6S4BwL9gQGSDq92XBcCj0ZE/4i4GrgZGJH27wVsHxHzaxmTscDZEdEfOAxYK+lzwNHAwRHRD/hpqlvXWHWJiMER8XPgGuDqiDgojcPNNXUcEeMjoiwiyujcrZbwzMzMmlfelzBeiIi5absCKCX7Iz4jld0K3FlQ/970vRJYGBEvAUhaAuxOlrj8JSLeSuV3kSUUbYB7ImJtKp+cvncCPg7cKamqj+0L+rs7IjYCT0vqnsqOSF9z0utOZMnRzDqO807gYkkXAKcBE+uoOwu4StIk4K6IeFHSUGBCRKwBiIjXJXWuZ6zuKNgeCuxbcIwfkFQSEW/WEYeZmVmryHsytK5gewPQpYH1N1bbdyPZWGqTPTK1lbcBVqZZmPriU8H3H0fEjfXE+p6IWCPpYeCLwAn898xU9bo/kTQF+DzwREqEBDT2ssO3CrbbAIOqksGGGLCmgvKK2oatyMb4Ckwzs21J3pfJqlsFrJB0WHp9CjCjjvrVzQSOlrSDpB2BY4BHgb8DwyV1SLNBwwAi4g3gBUnHAyjTr54+pgKnpXaQ9BFJH6pW502gpFrZzcC1wJMR8XptjUsqjYjKiBgHlAN9gIdSnzukOjtFRGPG6iFgdEEftSV/ZmZmrS7vM0M1ORW4If3hX0J2DlCDRMRTkiYC/0hFN0fEHABJ9wLzgH+RJRmrUp2TgOsl/Q/QHvhDqldbHw9J2gd4PC07rQZOBl4pqDYfWC9pHjAxIq6OiApJbwAT6jmM8yV9kmym7GnggYhYlxKYcknvAPcD36PhY3UucJ2k+WQ/czOBUfXEYWZm1ip808VWIqlTRKxOicNM4KyIeKoV+98VmA70SechbdHKdleUn19/vaLwMpmZbaHkmy42iZfJWs94SXOBp4A/t3Ii9DVgNnDR1pAImZmZtSYvk7WSiCjaTRAj4jayy+DfI2kkcF61qrMi4uxWC8zMzGwL4GUy2yKVlZVFebmfTWZm1hheJmsaL5OZmZlZrjkZMjMzs1xzMmRmZma55mTIzMzMcs3JkJmZmeWakyEzMzPLNSdDZmZmlmtOhszMzCzXnAyZmZlZrjkZMjMzs1xzMmRmZma55ge12hZpVWUlk0tLix2GWbMavnhxsUMwsxp4ZsjMzMxyzcmQmZmZ5ZqTITMzM8s1J0NmZmaWa06GzMzMLNcUEcWOwWwT6l0W3Fhe7DDMml0MKXYEti2TVBERZcWOY2uzVc4MSTpX0j8lTWri/kMk3Ze2L5E0tnkj3HyFcUmaKOm4YsdkZma2Ldpa7zP0TeBzEfFCsQOxuklqFxHrix2HmZlZbba6mSFJNwB7A/dKGiPpbknzJT0h6YBUp1JSF2Vek/S1VH67pKF1tH2mpAckdZRUKulBSRWSHpXUJ9U5XtICSfMkzayjrREptsmSXpA0WtK3Jc1Jse5U0OeTqb0/S9qhnuP/dGqjUtItkraXNFDSXen9L0paK2k7SR0kLUnl50p6Oo3VH+pof7CkuelrjqSSVP6d1Oc8ST9JZf3TscyX9BdJH0zl0yVdIWkGcJ6kbunYnkxfh9bS91mSyiWVs2p5XcNgZmbWbLa6ZCgiRgH/AT4J9ADmRMQBwPeA21K1WcChQF9gCXBYKj8EeKKmdiWNBoYDR0fEWmA8cE5EDADGAr9OVb8PfDYi+gFH1RPufsBXgYHA5cCaiDgQeBz4WqpzV0QclNr7J3B6bY1J6gBMBE6MiP3JZva+ATwFHJiqHQYsAA4CDgZmp/ILgQPTWI2qI+axwNkR0T+1tVbS54CjgYNTnD9NdW8DvpvarAR+UNBOl4gYHBE/B64Bro6Ig4BjgZtr6jgixkdEWUSU0blbHSGamZk1n611mazKJ8j+uBIRf5O0s6TOwKPA4cC/gOuBsyR9BHg9IlZLqt7OKcCLZInQu5I6AR8H7iyou336PguYKOmPwF31xPdIRLwJvClpFTA5lVcCB6Tt/ST9COgCdAKm1tFeb+CFiFiUXt9Klrj8QtLzkvYhS7yuSsffNo0FwHxgkqS7gbvr6GMWcFU6H+uuiHgxzaZNiIg1ABHxehrnLhExoyCWOwvauaNgeyiwb8FYfkBSSRobMzOzotrak6FNshoggJnA2cAewEXAMcBxvJ8YVLcA6A/sBrxANmO2Ms2O/HfjEaMkHQwMA+ZK6h8Rr9XS7rqC7Y0Frzfy/thPJEvC5kkaAQyppS2o+XirPAp8DngX+Gtqty3ZTA8p3sPJZrMultS3pnN5IuInkqYAnweeSImQyMa1Md4q2G4DDEozbg0yYE0F5RV1Ha5ZCxnjK2zN8marWyarZiZwEmRXiAGvRsQbEfF/QFegZ0QsAf5OlhTUlgzNAb5Odh7SrhHxBvCCpONT25LUL22XRsTsiPg+8Cqw+2YeQwnwkqT2VcdSh2eAHpI+ml6fAlTNzMwEzgcej4jlwM5AH2ChpDbA7hHxCPAd3p+F2kQ6vsqIGAeUpzYeAk6rOp9J0k4RsQpYIalqCbIwluoeAkYX9LFJkmlmZlYsW/vM0CXABEnzgTXAqQXvzSabGYEsCfoxWVJUo4j4u7JL2adI+gxZYnK9pP8B2gN/AOYBV0rqSTZbMi2VbY6LU6z/Ils+K6kjxrcljSRbvmsHPAnckN6eDXQnS4ogWxZ7JSIi1f1tWtoS2fk7K2vp5nxJnwQ2AE8DD0TEupTAlEt6B7if7BytU4EbUpK0BBhZS5vnAtelz6ldirGu85bMzMxajW+6aFukst0V5ecXOwrLJS+T2VZMvulik2zty2RmZmZmm2VrXyYrOkmfBcZVK34hIo4pRjwNlZbbzqtWPCsizi5GPGZmZsXiZTLbIpWVlUV5uZ9NZmbWGF4maxovk5mZmVmuORkyMzOzXHMyZGZmZrnmZMjMzMxyzcmQmZmZ5ZqTITMzM8s1J0NmZmaWa06GzMzMLNecDJmZmVmuORkyMzOzXHMyZGZmZrnmB7XaFmlVZSWTS0uLHYZZgwxfvLjYIZjZZvDMkJmZmeWakyEzMzPLNSdDZmZmlmtOhszMzCzXnAyZmZlZrikiih2D2SbUuyy4sbzYYZg1SgwpdgSWd5IqIqKs2HFsbTwz1AIkTZR0XD11RkjatbViMjMzs5o5GSqeEcA2nwxJalvsGMzMzOqSq2RI0tckzZc0T9LtkvaUNC2VTZO0R6o3UdL1kh6RtETSYEm3SPqnpIkF7a2W9HNJT6X9u9XQ5wBJMyRVSJoqaZc0a1QGTJI0V1LHmuql/adLGifpH5IWSToslbeVdKWkJ1P8X0/lu0iamdpdIOmwVHdiel0p6Vt1jNG5kp5Obf4hlXWSNCHtO1/Ssan8K6lsgaRx1cblMkmzgUG1HVsNfZ8lqVxSOauWN/bjNTMza5LcJEOS+gIXAZ+KiH7AecCvgNsi4gBgEnBtwS4fBD4FfAuYDFwN9AX2l9Q/1dkReCoiPgbMAH5Qrc/2wC+B4yJiAHALcHlE/AkoB06KiP7A+prqFTTVLiIGAucX9HE6sCoiDgIOAs6UtBfwVWBqarcfMBfoD3wkIvaLiP2BCXUM1YXAgWlMRqWyi1Nf+6fyv6UlvnFpjPoDB0k6umBcFkTEwcDseo7tPRExPiLKIqKMzpvklWZmZi0iT4/j+BTwp4h4FSAiXpc0CPhSev924KcF9SdHREiqBF6OiEoASQuBHmRJxkbgjlT/t8Bd1frsDewHPCwJoC3wUg2x1Vevqt2K1DfAEcABBecmdQZ6Ak8Ct6RE7O6ImCtpCbC3pF8CU4CHahkjgPlkM1Z3A3ensqHAl6sqRMQKSYcD0yNieRqXScDhaZ8NwJ8bOQZmZmZFkadkSEB9l84Vvr8ufd9YsF31urZxq96+gIURMagBsdVVr6r/DQV9CzgnIqZu0liWqAwDbpd0ZUTcJqkf8FngbOAE4LRa+hpGltQcBVycZtRqGjvVcTxvR8SGBh5bjQasqaC8oq4ubKs1xlewmtmWJTfLZMA04ARJOwNI2gl4jPdnPE4C/t7INtsAVTMzX61h/2eBbmkGCkntU3IB8CZQ0oB6tZkKfCPNACGpl6QdJe0JvBIRNwG/AT4mqSvQJiL+TLbk9bGaGpTUBtg9Ih4BvgN0ATqRzSSNLqj3QbLlr8GSuqaTpL9CtlRYXVOOzczMrNXkZmYoIhZKuhyYIWkDMAc4l2xJ6QJgOTCykc2+BfSVVAGsAk6s1uc7aRnrWkmdycb7F8BCYCJwg6S1wCCypKqmerW5mWzJ7Cll6y289doAAA/mSURBVE/LgaOBIcAFkt4FVgNfAz4CTEjJDsD/q6XNtsBvUwwCro6IlZJ+BFwnaQHZ7NSlEXGXpP8HPJLq3h8R91RvsJ4xMDMzKzrfdHEzSFodEZ2KHce2qGx3Rfn5xY7CWoSXycxajHzTxSbJ0zKZmZmZ2SZys0zWErbmWSFJ1wGHViu+JiLquuzezMxsm+NkKKci4uxix1Cn7gNgjJ9NZmZmLc/LZGZmZpZrTobMzMws15wMmZmZWa45GTIzM7NcczJkZmZmueZkyMzMzHLNyZCZmZnlmpMhMzMzyzUnQ2ZmZpZrTobMzMws15wMmZmZWa752WS2RVpVWcnk0tJih1EUwxcvLnYIZma54pkhMzMzyzUnQ2ZmZpZrTobMzMws15wMmZmZWa45GTIzM7NcU0QUOwazTah3WXBjebHDKLoYUuwIzGxrIqkiIsqKHcfWxjNDgKQRkn6VtkdJ+loT2ugi6ZtN2O8SSWMbu19TtXZ/ZmZmWzonQ9VExA0RcVsTdu0CNDoZagxJW919oSS1LXYMZmZmdclFMiTpbkkVkhZKOiuVjZS0SNIM4NCCuu/NnEiaLqksbXeVtDRt95X0D0lzJc2X1BP4CVCayq5M9S6Q9GSqc2lBHxdJelbSX4He9cQ+XdIVKc7zJA2XNFvSHEl/ldS9IO5bUv0lks5tYn/nSno6xfyHVNZJ0gRJlan82FT+lVS2QNK4gjZWS7pM0mxgkKQBkmakz2CqpF1q6fssSeWSylm1vK4wzczMms1WN9PQRKdFxOuSOgJPSpoCXAoMAFYBjwBzGtHeKOCaiJgkaTugLXAhsF9E9AeQdATQExgICLhX0uHAW8CXgQPJxv8poKKe/rpExODU7geBQyIiJJ0BfAcYk+r1AT4JlADPSroeOKCR/V0I7BUR6yR1SWUXA6siYv+qGCTtCowjG8MVwEOSjo6Iu4EdgQUR8X1J7YEZwBcjYrmkE4HLgdOqdxwR44HxkM4ZMjMzawV5SYbOlXRM2t4dOAWYHhHLASTdAfRqRHuPAxdJ2g24KyKek1S9zhHpqyrJ6kSWHJUAf4mINanvexvQ3x0F27sBd6TZle2AFwremxIR64B1kl4BugOHNbK/+cAkSXcDd6eyoWQJFQARsSIldoVjOAk4PO2zAfhzqt4b2A94OI1RW+ClBhyzmZlZq9jmkyFJQ8j+mA+KiDWSpgPPAPs0YPf1vL+U2KGqMCJ+l5aAhgFT0wzNkupdAz+OiBurxXM+0NhZj7cKtn8JXBUR96Zju6TgvXUF2xt4//NtTH/DyJKao4CLJfUlO5bqbWyS/RV4OyI2FNRbGBGDGhEDA9ZUUF6RuhjjSSIzM2s5eThnqDOwIiVCfYBDgI7AEEk7p2Wc42vZdynZMhDAcVWFkvYGlkTEtcC9ZEtRb5LN+lSZCpwmqVPa5yOSPgTMBI6R1FFSCTC8Ccfz77R9agPqN7g/SW2A3SPiEbLlty5kM1oPAaML6n0QmA0MTudStQW+QrYcVt2zQDdJg9K+7VOCZWZmtkXY5meGgAeBUZLmk/1hfoJsmeYSsuWul8jOoym86qlqKuJnwB8lnQL8reD9E4GTJb0LLAMuS+ckzZK0AHggIi6QtA/weFoeWg2cHBFPpWW5ucC/gEcbeTyXAHdK+nc6lr3qqtzI/toCv5XUmWxG5+qIWCnpR8B16dg2AJdGxF2S/h/Z+VYC7o+Ie2ro/x1JxwHXpnbbAb8AFjbqqM3MzFqIb7pYjaRfAk9FxIRix5JnZbsrys9PL7xMZmbWIPJNF5skD8tkDSbph8DBZEtfZmZmlgN5WCZrsIi4mOwy8lYn6ToK7neUXNNSM1St3Z+ZmdmWystktkUqKyuL8nI/m8zMrDG8TNY0XiYzMzOzXHMyZGZmZrnmZMjMzMxyzcmQmZmZ5ZqTITMzM8s1J0NmZmaWa06GzMzMLNecDJmZmVmuORkyMzOzXHMyZGZmZrnmZMjMzMxyzQ9qtS3SqspKJpeWFjsMa2HDFy8udghmZp4ZMjMzs3xzMmRmZma55mTIzMzMcs3JkJmZmeWakyEzMzPLNUVEsWMw24R6lwU3lhc7DGtFMaTYEZht/SRVRERZsePY2hRlZkjSCEm/StujJH2tCW10kfTNJux3iaSxjd2vEe2PkLTrZuw/RNJ9dbx/lKQLG9DOlZIWSrqyqbGYmZnlQdHvMxQRNzRx1y7AN4FfN2M4zWEEsAD4T0s0HhH3Avc2oOrXgW4Rsa4l4mgoSW0jYkMxYzAzM6tLi8wMSbpbUkWamTgrlY2UtEjSDODQgrrvzdRImi6pLG13lbQ0bfeV9A9JcyXNl9QT+AlQmsquTPUukPRkqnNpQR8XSXpW0l+B3vXEPl3S1ZJmSvqnpIMk3SXpOUk/SnV6SFpQsM/YdBzHAWXApBRXR0lLJXVN9cokTU/bAyU9JmlO+l5nXAV9Fc6qTZR0bdp/SeofSfcCOwKzJZ0oaU9J09K4TJO0Rx3tHy9pgaR5kmamsraSfiapMrVxTir/dIq/UtItkrZP5UslfV/S34HjJZVKejD9TDwqqU8tfZ8lqVxSOauWN2Q4zMzMNltLzQydFhGvS+oIPClpCnApMABYBTwCzGlEe6OAayJikqTtgLbAhcB+EdEfQNIRQE9gICDgXkmHA28BXwYOJDvep4CKevp7JyIOl3QecE+K+3VgsaSra9spIv4kaTQwNiLKU1y1VX8GODwi1ksaClwBHFtPXDXZBfgE0IdsxuhPEXGUpNUFYzMZuC0ibpV0GnAtcHQt7X0f+GxE/FtSl1R2FrAXcGCKdydJHYCJwKcjYpGk24BvAL9I+7wdEZ9I/U8DRkXEc5IOJpvN+1T1jiNiPDAe0jlDZmZmraClkqFzJR2TtncHTgGmR8RyAEl3AL0a0d7jwEWSdgPuSn9Uq9c5In1VJVmdyJKjEuAvEbEm9d2QJaaqOpXAwoh4Ke27JB3PykbEXpvOwK1pliuA9k1s5+6I2Ag8Lal7LXUGAV9K27cDP62jvVnAREl/BO5KZUOBGyJiPUBKdPsBL0TEolTnVuBs3k+G7gCQ1An4OHBnwWe2fSOOz8zMrEU1ezIkaQjZH89BEbEmLQs9A+zTgN3X8/7SXYeqwoj4naTZwDBgqqQzgCXVuwZ+HBE3VovnfLJkozGqzrPZWLBd9bpdtTj/K9Ya1HhMwA+BRyLiGEk9gOmNjLF6rJCNQUPUOh4RMSrN3gwD5krqn9qtvk99fb2VvrcBVlbNUjXUgDUVlFc09HC2cmM8CWZmVkwtcc5QZ2BFSoT6AIcAHYEhknaW1B44vpZ9l5ItSQEcV1UoaW9gSURcSzZrcwDwJtmsT5WpwGlpJgJJH5H0IWAmcEw6f6cEGN4Mx/gy8KF0PNsDXyh4r3pchcdUuAzWGfh32h7RDDHV5TGypUKAk4C/11ZRUmlEzI6I7wOvks2EPQSMktQu1dmJLMHtIemjaddTgBnV24uIN4AXJB2f9lWaVTIzM9sitEQy9CDQTtJ8stmPJ4CXgEvIlrv+SnbeTqGq/xr/DPiGpMeArgXvnwgskDSX7NyY2yLiNWBWOtn3yoh4CPgd8LikSuBPQElEPEW2ZDMX+DPw6OYeYES8C1wGzAbuI0sMqkwEbqg6gZrsXKlrJD0KFF5V9VPgx5JmkZ0D1ZLOBUamz+QU4Lw66l6ZToheQJZIzgNuBv4XmC9pHvDViHgbGEm2/FVJNmtW25WBJwGnp30XAl9sjoMyMzNrDkW/6aKkXwJPRcSEogZiW5Sy3RXl5xc7ilbiZTIzaybyTRebpKiP45D0Q+BgGnbfHDMzM7NmV9SbLkbExcDFxehb0nUU3O8ouWZLmaGSNJJNl7NmRcTZzdT+RWx67tadEXF5c7RvZma2tSj6MplZTcrKyqK83M8mMzNrDC+TNY2fWm9mZma55mTIzMzMcs3JkJmZmeWakyEzMzPLNSdDZmZmlmtOhszMzCzXnAyZmZlZrjkZMjMzs1xzMmRmZma55jtQ2xZJ0pvAs8WOo8i6Aq8WO4gi8xh4DMBjAA0fgz0joltLB7OtKeqzyczq8Gzebykvqdxj4DHwGHgMwGPQ0rxMZmZmZrnmZMjMzMxyzcmQbanGFzuALYDHwGMAHgPwGIDHoEX5BGozMzPLNc8MmZmZWa45GTIzM7NcczJkRSXpSEnPSnpe0oU1vL+9pDvS+7Ml9Wj9KFtWA8bg25KeljRf0jRJexYjzpZU3xgU1DtOUkja5i4xbsgYSDoh/SwslPS71o6xpTXgd2EPSY9ImpN+Hz5fjDhbiqRbJL0iaUEt70vStWl85kv6WGvHuM2KCH/5qyhfQFtgMbA3sB0wD9i3Wp1vAjek7S8DdxQ77iKMwSeBHdL2N/I4BqleCTATeAIoK3bcRfg56AnMAT6YXn+o2HEXYQzGA99I2/sCS4sddzOPweHAx4AFtbz/eeABQMAhwOxix7ytfHlmyIppIPB8RCyJiHeAPwBfrFbni8CtaftPwKclqRVjbGn1jkFEPBIRa9LLJ4DdWjnGltaQnwOAHwI/Bd5uzeBaSUPG4EzguohYARARr7RyjC2tIWMQwAfSdmfgP60YX4uLiJnA63VU+SJwW2SeALpI2qV1otu2ORmyYvoI8H8Fr19MZTXWiYj1wCpg51aJrnU0ZAwKnU72P8NtSb1jIOlAYPeIuK81A2tFDfk56AX0kjRL0hOSjmy16FpHQ8bgEuBkSS8C9wPntE5oW4zG/nthDeTHcVgx1TTDU/1eDw2pszVr8PFJOhkoAwa3aEStr84xkNQGuBoY0VoBFUFDfg7akS2VDSGbHXxU0n4RsbKFY2stDRmDrwATI+LnkgYBt6cx2Njy4W0RtvV/D4vGM0NWTC8Cuxe83o1Np73fqyOpHdnUeF3TyFubhowBkoYCFwFHRcS6VoqttdQ3BiXAfsB0SUvJzpW4dxs7ibqhvwv3RMS7EfEC2YOMe7ZSfK2hIWNwOvBHgIh4HOhA9gDTvGjQvxfWeE6GrJieBHpK2kvSdmQnSN9brc69wKlp+zjgb5HOJNxG1DsGaYnoRrJEaFs7TwTqGYOIWBURXSOiR0T0IDtv6qiIKC9OuC2iIb8Ld5OdTI+krmTLZktaNcqW1ZAx+F/g0wCS9iFLhpa3apTFdS/wtXRV2SHAqoh4qdhBbQu8TGZFExHrJY0GppJdSXJLRCyUdBlQHhH3Ar8hmwp/nmxG6MvFi7j5NXAMrgQ6AXemc8f/NyKOKlrQzayBY7BNa+AYTAWOkPQ0sAG4ICJeK17UzauBYzAGuEnSt8iWh0ZsS/85kvR7smXQrum8qB8A7QEi4gay86Q+DzwPrAFGFifSbY8fx2FmZma55mUyMzMzyzUnQ2ZmZpZrTobMzMws15wMmZmZWa45GTIzM7NcczJkZmZmueZkyMzMzHLt/wON0CQW+n8KjAAAAABJRU5ErkJggg==\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "#Transformation family 2\n",
    "# D = (max(S) / (1 + S - min(S))) ** \n",
    "# D = (max(S) / (1 + S - min(S))) ** 2\n",
    "# D = np.log((max(S) / (1 + S - min(S))))\n",
    "\n",
    "#Please add your code here\n",
    "## Calculation of different dissimilarity matrices with different transformations of the similarity matrix\n",
    "S  = prodom[\"similarity\"]\n",
    "D1 = ( np.max(S) / (1 + S - np.min(S)) ) ** 1\n",
    "D2 = ( np.max(S) / (1 + S - np.min(S)) ) ** 0.5\n",
    "D3 = np.log( np.max(S) / (1 + S - np.min(S)) )\n",
    "D  = [D1, D2, D3]\n",
    "\n",
    "# Calculation of different cluster predictions with different dissimilarity matrices\n",
    "labels2_SC = []\n",
    "for i in range(3):\n",
    "    SC = skl.cluster.SpectralClustering(n_clusters=4, random_state=42, affinity=\"precomputed\")\n",
    "    affinities = np.exp(- D[i] / 3) # SC takes positive affinities as input\n",
    "    labels2_SC.append(SC.fit_predict(affinities))\n",
    "    \n",
    "# Print and plot the scores for different functions and different params\n",
    "best_transform2_SC = compare_clusterings(labels_true=prodom[\"labels\"],\n",
    "                                         width = 0.2,\n",
    "                                         SpectralClustering_a1=labels2_SC[0],\n",
    "                                         SpectralClustering_a1_2=labels2_SC[1],\n",
    "                                         SpectralClustering_alog=labels2_SC[2])\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "colab_type": "text",
    "id": "ZcMJIbYGNHfY"
   },
   "source": [
    "<p style=\"background-color:#adebad;\">\n",
    "3. Provide a description of another reasonable transformation from similarity to dissimilarity and explain your choice.\n",
    "</p>"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "metadata": {},
   "outputs": [],
   "source": [
    "## Convert best transform parameter to the index\n",
    "def paramToIndex(param):\n",
    "    if param==\"a=1\":\n",
    "        return 0\n",
    "    elif param==\"a=1/2\" or param==\"a=2\":\n",
    "        return 1\n",
    "    elif param==\"a=exp\" or param==\"a=log\":\n",
    "        return 2"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Best of Transform 1 has the highest score for 0 different functions.\n",
      "Best of Transform 2 has the highest score for 7 different functions.\n",
      "The winner is Best of Transform 2.\n"
     ]
    },
    {
     "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>Best of Transform 1</th>\n",
       "      <th>Best of Transform 2</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>v_measure_score</th>\n",
       "      <td>0.210681</td>\n",
       "      <td>0.385255</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>normalized_mutual_info_score</th>\n",
       "      <td>0.210681</td>\n",
       "      <td>0.385255</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>homogeneity_score</th>\n",
       "      <td>0.144479</td>\n",
       "      <td>0.301001</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>fowlkes_mallows_score</th>\n",
       "      <td>0.525048</td>\n",
       "      <td>0.562150</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>completeness_score</th>\n",
       "      <td>0.388861</td>\n",
       "      <td>0.535012</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>adjusted_rand_score</th>\n",
       "      <td>0.061500</td>\n",
       "      <td>0.202437</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>adjusted_mutual_info_score</th>\n",
       "      <td>0.209615</td>\n",
       "      <td>0.384531</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                              Best of Transform 1  Best of Transform 2\n",
       "v_measure_score                          0.210681             0.385255\n",
       "normalized_mutual_info_score             0.210681             0.385255\n",
       "homogeneity_score                        0.144479             0.301001\n",
       "fowlkes_mallows_score                    0.525048             0.562150\n",
       "completeness_score                       0.388861             0.535012\n",
       "adjusted_rand_score                      0.061500             0.202437\n",
       "adjusted_mutual_info_score               0.209615             0.384531"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "###### SPECTRAL CLUSTERING\n",
    "## Comparison of the tranforms above using their best parameters\n",
    "labels_transform1best_SC = labels1_SC[paramToIndex(best_transform1_SC)]\n",
    "labels_transform2best_SC = labels2_SC[paramToIndex(best_transform2_SC)]\n",
    "\n",
    "_ = compare_clusterings(labels_true=prodom[\"labels\"],\n",
    "                        width = 0.2,\n",
    "                        SpectralClustering_aTransform1=labels_transform1best_SC,\n",
    "                        SpectralClustering_aTransform2=labels_transform2best_SC)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<p style=\"background-color:#adebad;\">\n",
    "4. Implement your transformation. <br>\n",
    "5. Compare your transformation using <b>compare_clusterings()</b> with the best from task 4.0.2. You should get one plot. <br>\n",
    "</p>"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "a=1 has the highest score for 0 different functions.\n",
      "a=2 has the highest score for 0 different functions.\n",
      "a=exp has the highest score for 7 different functions.\n",
      "The winner is a=exp.\n"
     ]
    },
    {
     "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>a=1</th>\n",
       "      <th>a=2</th>\n",
       "      <th>a=exp</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>v_measure_score</th>\n",
       "      <td>0.208276</td>\n",
       "      <td>0.126268</td>\n",
       "      <td>0.232029</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>normalized_mutual_info_score</th>\n",
       "      <td>0.208276</td>\n",
       "      <td>0.126268</td>\n",
       "      <td>0.232029</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>homogeneity_score</th>\n",
       "      <td>0.142113</td>\n",
       "      <td>0.096585</td>\n",
       "      <td>0.156427</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>fowlkes_mallows_score</th>\n",
       "      <td>0.524901</td>\n",
       "      <td>0.455521</td>\n",
       "      <td>0.532157</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>completeness_score</th>\n",
       "      <td>0.389709</td>\n",
       "      <td>0.182291</td>\n",
       "      <td>0.449060</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>adjusted_rand_score</th>\n",
       "      <td>0.086055</td>\n",
       "      <td>0.080498</td>\n",
       "      <td>0.088868</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>adjusted_mutual_info_score</th>\n",
       "      <td>0.207743</td>\n",
       "      <td>0.125743</td>\n",
       "      <td>0.231505</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                                   a=1       a=2     a=exp\n",
       "v_measure_score               0.208276  0.126268  0.232029\n",
       "normalized_mutual_info_score  0.208276  0.126268  0.232029\n",
       "homogeneity_score             0.142113  0.096585  0.156427\n",
       "fowlkes_mallows_score         0.524901  0.455521  0.532157\n",
       "completeness_score            0.389709  0.182291  0.449060\n",
       "adjusted_rand_score           0.086055  0.080498  0.088868\n",
       "adjusted_mutual_info_score    0.207743  0.125743  0.231505"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "#Please add your code here\n",
    "###### PAIRWISE CLUSTERING\n",
    "def sigmoid(x):\n",
    "    return 1 / (1 + np.exp(-x))\n",
    "\n",
    "# D = (sigmoid(-(S-min(S))/(max(S)-min(S)))) ** a\n",
    "# a = [1, 2, exp]\n",
    "\n",
    "## Calculation of different dissimilarity matrices with different transformations of the similarity matrix\n",
    "S  = prodom[\"similarity\"]\n",
    "\n",
    "D1 = (sigmoid(-(S-np.min(S))/(np.max(S)-np.min(S)))) ** 1\n",
    "D2 = (sigmoid(-(S-np.min(S))/(np.max(S)-np.min(S)))) ** 2\n",
    "D3 = np.exp(sigmoid(-(S-np.min(S))/(np.max(S)-np.min(S))))\n",
    "D  = [D1, D2, D3]\n",
    "\n",
    "# Calculation of different cluster predictions with different dissimilarity matrices\n",
    "labels3_PC = []\n",
    "for i in range(3):\n",
    "    PC = PairwiseClustering()\n",
    "    labels3_PC.append(PC.fit_predict(D[i]))\n",
    "    \n",
    "best_transform3_PC = compare_clusterings(labels_true=prodom[\"labels\"],\n",
    "                                         width = 0.2,\n",
    "                                         PairwiseClustering_a1   = labels3_PC[0], \n",
    "                                         PairwiseClustering_a2   = labels3_PC[1],\n",
    "                                         PairwiseClustering_aexp = labels3_PC[2])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 30,
   "metadata": {},
   "outputs": [
    {
     "ename": "NameError",
     "evalue": "name 'labels1_PC' is not defined",
     "output_type": "error",
     "traceback": [
      "\u001b[0;31m---------------------------------------------------------------------------\u001b[0m",
      "\u001b[0;31mNameError\u001b[0m                                 Traceback (most recent call last)",
      "\u001b[0;32m<ipython-input-30-9f6f2e03f85b>\u001b[0m in \u001b[0;36m<module>\u001b[0;34m\u001b[0m\n\u001b[1;32m      1\u001b[0m \u001b[0;31m###### PAIRWISE CLUSTERING\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m      2\u001b[0m \u001b[0;31m## Comparison of the given transforms and the additionally implemented transform using their best parameters\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m----> 3\u001b[0;31m \u001b[0mlabels_transform1best_PC\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mlabels1_PC\u001b[0m\u001b[0;34m[\u001b[0m\u001b[0mparamToIndex\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mbest_transform1_PC\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m]\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m      4\u001b[0m \u001b[0mlabels_transform2best_PC\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mlabels2_PC\u001b[0m\u001b[0;34m[\u001b[0m\u001b[0mparamToIndex\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mbest_transform2_PC\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m]\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m      5\u001b[0m \u001b[0mlabels_transform3best_PC\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mlabels3_PC\u001b[0m\u001b[0;34m[\u001b[0m\u001b[0mparamToIndex\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mbest_transform3_PC\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m]\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;31mNameError\u001b[0m: name 'labels1_PC' is not defined"
     ]
    }
   ],
   "source": [
    "###### PAIRWISE CLUSTERING\n",
    "## Comparison of the given transforms and the additionally implemented transform using their best parameters\n",
    "labels_transform1best_PC = labels1_PC[paramToIndex(best_transform1_PC)]\n",
    "labels_transform2best_PC = labels2_PC[paramToIndex(best_transform2_PC)]\n",
    "labels_transform3best_PC = labels3_PC[paramToIndex(best_transform3_PC)]\n",
    "\n",
    "_ = compare_clusterings(labels_true=prodom[\"labels\"],\n",
    "                        width = 0.2,\n",
    "                        PairwiseClustering_aTransform1=labels_transform1best_PC,\n",
    "                        PairwiseClustering_aTransform2=labels_transform2best_PC,\n",
    "                        PairwiseClustering_aTransform3=labels_transform3best_PC)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "colab_type": "text",
    "id": "MzDZ4INKNHfg"
   },
   "source": [
    "<h2 style=\"background-color:#f0b375;\">\n",
    "Section 5.5 \n",
    "<span style=font-size:50%> Complete all problems in this (5.5.1-5.5.6) and previous sections to get a grade of 5.5 </span>\n",
    "</h2>"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "In this section, you have to  implement and calculate cost function and investigate the stability of pairwise clustering in terms of obtained values of the clustering cost function, providing experiments for different random seeds and number of clusters."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<p style=\"background-color:#adebad;\">\n",
    "1. Implement the cost function of the <b>PairwiseClustering class</b>. <br>\n",
    "2. Compute the cost function for <b>Protein</b> data. <br>\n",
    "3. Calculate the execution time. <br>\n",
    "</p>"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 31,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 51
    },
    "colab_type": "code",
    "id": "-X4_uD43NHfh",
    "outputId": "1e9a2720-7aa5-47d0-8895-ada2067923d1"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "CPU times: user 1.09 s, sys: 35.3 ms, total: 1.13 s\n",
      "Wall time: 311 ms\n"
     ]
    }
   ],
   "source": [
    "%%time\n",
    "PC = PairwiseClustering()\n",
    "labels_pair = PC.fit_predict(protein[\"dissimilarity\"])\n",
    "cost = PC.cost(protein[\"dissimilarity\"], PC.M)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "colab_type": "text",
    "id": "1kTYZi6wNHfo"
   },
   "source": [
    "<p style=\"background-color:#adebad;\">\n",
    "4. Re-run <b>fit_predict()</b> with 10 different random seeds and plot the distribution of clustering costs as shown in Figure 1 of reference [1]. <br>\n",
    "5. Visualize the histograms of obtained cost functions.You should get 2 histograms; one for <b>Prodom</b> and the other for <b>Protein</b> data. Hint: use <b>matplotlib.plot.hist()</b> to plot histograms of cost functions.\n",
    "</p>"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<p style=\"background-color:#adebad;\">\n",
    "<b>NOTE:</b> For ProDom data use the best transformation from section 5.0 to transform similiarities to dissimiliraties. \n",
    "</p>"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 34,
   "metadata": {
    "colab": {},
    "colab_type": "code",
    "id": "O0JuhulcB0CE"
   },
   "outputs": [],
   "source": [
    "# data = [protein, prodom]\n",
    "# Use the best transform from Problem 3 to transform similarities\n",
    "# other parameters are default\n",
    "S_prodom  = prodom[\"similarity\"]\n",
    "D_prodom  = np.log( np.max(S) / (1 + S - np.min(S)) )\n",
    "D_protein = protein[\"dissimilarity\"]\n",
    "n_seeds   = 10\n",
    "data      = [D_protein, D_prodom]\n",
    "costs1    = np.zeros((len(data), n_seeds))\n",
    "for j in range(len(data)):\n",
    "    for i in range(n_seeds):\n",
    "        PC = PairwiseClustering(random_state=i)\n",
    "        PC.fit_predict(data[j])\n",
    "        costs1[j, i] = PC.cost(data[j], PC.M)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<p style=\"background-color:#adebad;\">\n",
    "6. Perform the experiment for all the settings listed below, and visualize the results properly. You should get 4 histograms; one for each different number of clusters. <br>\n",
    "</p>"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 36,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 576x432 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 576x432 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "ds  = [\"Protein\", \"Prodom\"]\n",
    "for i in range(costs1.shape[0]):\n",
    "    fig = plt.figure(figsize=(8,6))\n",
    "    plt.hist(costs1[i, :], label=ds[i])\n",
    "    plt.legend(bbox_to_anchor=(1, 1), loc='upper left', ncol=1)\n",
    "    plt.title(\"Histogram of the Costs over 10 Different Random Seeds for \" + ds[i] + \" Dataset\")\n",
    "    plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "colab_type": "text",
    "id": "QJpukq-eNHf8"
   },
   "source": [
    "<h2 style=\"background-color:#f0b375;\">\n",
    "Section 6.0\n",
    "<span style=font-size:50%> Complete all problems (6.0.1-6.0.5) in this and previous sections to get a grade of 6.0 </span>\n",
    "</h2>\n",
    "\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "In this section, you must implement an EM-like procedure to obtain data embeddings in the Euclidean space, while visualizing these clusters on a 2d plane."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<p style=\"background-color:#adebad;\">\n",
    "1. Provide a description of the EM-like procedure for pairwise clustering in case we want to get embeddings (algorithm III in reference [1]). <br>\n",
    "</p>"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "#Provide your answer here"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<p style=\"background-color:#adebad;\">\n",
    "2.  Implement Algorithm III in [1], which explains how to compute an embedding within the EM-like procedure. In code <b>fit_predict()</b>: the <b>embed == True</b> part.<br>\n",
    "3. Predict a clustering for the <b>Protein</b> data.<br>\n",
    "4. Calculate the execution time.<br>\n",
    "</p>"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "colab": {},
    "colab_type": "code",
    "id": "wYurFjMcNHf8"
   },
   "outputs": [],
   "source": [
    "np.random.seed(111)\n",
    "#PCE = PairwiseClustering(embed=True,...)\n",
    "\n",
    "#Please add your code here"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "colab_type": "text",
    "id": "SiTEUrmANHf-"
   },
   "source": [
    "<p style=\"background-color:#adebad;\">\n",
    "5. Cluster the Protein data and properly visualize the embedding for <b>n_components=2</b>. Hint: <b>use matplotlib.pyplot.scatter()</b> to plot clusters on 2d plane using different colors for different clusters.\n",
    "</p>"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "#Please add your code here"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "colab_type": "text",
    "id": "MarVzEgQNHgB"
   },
   "source": [
    "## Comments\n",
    "\n",
    "We hope you found this exercise instructive.\n",
    "\n",
    "Feel free to leave comments below, we will read them carefully."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "colab_type": "text",
    "id": "5vKwjuCxNHgB"
   },
   "source": []
  }
 ],
 "metadata": {
  "accelerator": "GPU",
  "colab": {
   "name": "slt_ce_6.ipynb",
   "provenance": [],
   "toc_visible": true,
   "version": "0.3.2"
  },
  "kernelspec": {
   "display_name": "Python 3",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.7.6"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 1
}