4-Hausdorff_Symmetric_Distances-Medoids-Classification.ipynb

{"nbformat":4,"nbformat_minor":0,"metadata":{"colab":{"provenance":[],"toc_visible":true,"machine_shape":"hm","gpuType":"T4","collapsed_sections":["hYmigydfRqxk","BbN1UJIxRuNd"],"authorship_tag":"ABX9TyNyq2n5JWRNR6jpKQmL/ET5"},"kernelspec":{"name":"python3","display_name":"Python 3"},"language_info":{"name":"python"},"gpuClass":"standard","accelerator":"GPU"},"cells":[{"cell_type":"markdown","source":["# Installs + Imports"],"metadata":{"id":"hYmigydfRqxk"}},{"cell_type":"code","source":["# !pip install kmedoids\n","!pip install scikit-learn-extra"],"metadata":{"colab":{"base_uri":"https://localhost:8080/"},"id":"QrAthcf4eLEY","executionInfo":{"status":"ok","timestamp":1684134727774,"user_tz":-180,"elapsed":6170,"user":{"displayName":"Achil Pal","userId":"04608072963754583562"}},"outputId":"1d1b5eea-7b5c-4c4d-a2e2-be0b76a62635"},"execution_count":null,"outputs":[{"output_type":"stream","name":"stdout","text":["Looking in indexes: https://pypi.org/simple, https://us-python.pkg.dev/colab-wheels/public/simple/\n","Collecting scikit-learn-extra\n","  Downloading scikit_learn_extra-0.3.0-cp310-cp310-manylinux_2_17_x86_64.manylinux2014_x86_64.whl (2.0 MB)\n","\u001b[2K     \u001b[90m━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━\u001b[0m \u001b[32m2.0/2.0 MB\u001b[0m \u001b[31m65.3 MB/s\u001b[0m eta \u001b[36m0:00:00\u001b[0m\n","\u001b[?25hRequirement already satisfied: numpy>=1.13.3 in /usr/local/lib/python3.10/dist-packages (from scikit-learn-extra) (1.22.4)\n","Requirement already satisfied: scipy>=0.19.1 in /usr/local/lib/python3.10/dist-packages (from scikit-learn-extra) (1.10.1)\n","Requirement already satisfied: scikit-learn>=0.23.0 in /usr/local/lib/python3.10/dist-packages (from scikit-learn-extra) (1.2.2)\n","Requirement already satisfied: joblib>=1.1.1 in /usr/local/lib/python3.10/dist-packages (from scikit-learn>=0.23.0->scikit-learn-extra) (1.2.0)\n","Requirement already satisfied: threadpoolctl>=2.0.0 in /usr/local/lib/python3.10/dist-packages (from scikit-learn>=0.23.0->scikit-learn-extra) (3.1.0)\n","Installing collected packages: scikit-learn-extra\n","Successfully installed scikit-learn-extra-0.3.0\n"]}]},{"cell_type":"code","execution_count":null,"metadata":{"id":"gHRO_gVpRPgC"},"outputs":[],"source":["import numpy as np\n","import pandas as pd\n","import zipfile\n","import re\n","import warnings\n","import matplotlib.pyplot as plt\n","from sklearn_extra.cluster import KMedoids \n","from tensorflow.keras.utils import pad_sequences\n","from scipy.spatial.distance import directed_hausdorff\n","from itertools import combinations\n","from random import sample\n","from datetime import datetime\n","from sklearn.metrics import classification_report, zero_one_loss, coverage_error, silhouette_score, \\\n","                            label_ranking_loss, label_ranking_average_precision_score\n","from sklearn.svm import SVC\n","from sklearn.ensemble import RandomForestClassifier\n","from sklearn.linear_model import LogisticRegression\n","from sklearn.naive_bayes import MultinomialNB\n","warnings.filterwarnings(\"ignore\")"]},{"cell_type":"markdown","source":["# Helpers\n","\n"],"metadata":{"id":"BbN1UJIxRuNd"}},{"cell_type":"code","source":["def create_bags_of_sentences(docs_path, labels_path):\n","\n","  # Get the most frequent class only.\n","  labels = pd.read_csv(labels_path, header=None)\n","  labels = labels[0].map(lambda x: np.array([int(label) for label in x.split()]))\n","  labels = np.array(labels.tolist())\n","  most_frequent_counts = np.sum(np.transpose(labels), axis=1)\n","  most_frequent_index = most_frequent_counts.argmax()\n","  labels = labels[:, most_frequent_index]\n","\n","  # Open documents file.\n","  documents_file = open(docs_path, 'r')\n","\n","  # Initialize counters and bag of sentences dictionary.\n","  doc_counter = 0\n","  sentence_count = 0\n","  bag_of_sentences = {}\n","\n","  for document, label in zip(documents_file, labels):\n","    # Parse document.\n","    parsed_document = re.split(r'<\\d+>', document)\n","\n","    for sentence in parsed_document:\n","      # Remove leading and trailing whitespaces.\n","      sentence = sentence.strip()\n","\n","      # If sentence is not empty.\n","      if sentence:\n","        # Store words to an array of ints, since they are just ids.\n","        words = np.asarray(sentence.split(\" \"), dtype=np.int32)\n","        # Add a sentence to the bag.\n","        bag_of_sentences[sentence_count] = (doc_counter, words, label)\n","        sentence_count += 1\n","\n","    doc_counter += 1\n","\n","  # Close documents file.\n","  documents_file.close()\n","\n","  # Create dataframe of the bag.\n","  df = pd.DataFrame.from_dict(bag_of_sentences, orient='index',\n","                              columns = ['Bag', 'Sentence', 'Class'])\n","\n","  return df\n","\n","\n","def create_bag_per_document(df, max_words):\n","  ids, X, y = np.array(df['Bag']), np.array(df['Sentence']), np.array(df['Class'])\n","  N_unique_ids = np.unique(ids).shape[0]\n","  data = []\n","  labels = []\n","\n","  for i in range(N_unique_ids):\n","    bag = X[np.where(ids == i)]\n","    data.append(bag)\n","    label = y[np.where(ids == i)]\n","    labels.append(label)\n","\n","  data = np.array(data)\n","  labels = np.array(labels)\n","  labels = np.array([labels[i][0] for i in range(labels.shape[0])])\n","\n","  # Pad sentences for 'max_words' words per sentence max.\n","  for i, sentence in enumerate(data):\n","    data[i] = pad_sequences(sentence, maxlen=max_words, dtype='float32', padding='post')\n","\n","  return data, labels\n","\n","\n","def medoids_grid_search_and_cv(dist_matrix, K_list, cv=4):\n","  mean_scores = []\n","  for K in K_list:\n","    sil_list = []\n","    for _ in range(cv):\n","      # init_indices = sample(range(dist_matrix.shape[0]), K)\n","      model = KMedoids(n_clusters=K, random_state=1)\n","      preds = model.fit_predict(dist_matrix)\n","      sil_list.append(silhouette_score(dist_matrix, preds))\n","\n","    mean_scores.append(np.round(np.mean(sil_list), 4))\n","\n","  best_score = np.max(mean_scores)\n","\n","  print(f'Best Silhouette score found after Grid search : {best_score:.4f}')\n","  print(f'Best parameter : {K_list[np.argmax(mean_scores)]}')\n","\n","  return mean_scores\n","\n","\n","def transform_bags(X_train, train_medoids, X_test=None):\n","  if X_test is None:\n","    distances = np.zeros((X_train.shape[0], len(train_medoids)))    # X_transformed\n","    for i, x in enumerate(X_train):\n","      x = x.reshape(1, -1)\n","      for j, medoid_idx in enumerate(train_medoids):\n","        medoid = X_train[medoid_idx].reshape(1, -1)\n","        distances[i, j] = max(directed_hausdorff(x, medoid)[0], directed_hausdorff(medoid, x)[0])\n","  else:\n","    distances = np.zeros((X_test.shape[0], len(train_medoids)))    # X_transformed\n","    for i, x in enumerate(X_test):\n","      x = x.reshape(1, -1)\n","      for j, medoid_idx in enumerate(train_medoids):\n","        medoid = X_train[medoid_idx].reshape(1, -1)\n","        distances[i, j] = max(directed_hausdorff(x, medoid)[0], directed_hausdorff(medoid, x)[0])\n","  return distances"],"metadata":{"id":"Kxx7__tYRxmB"},"execution_count":null,"outputs":[]},{"cell_type":"code","source":["def create_one_bag_per_sentence(df):\n","  doc_dict = dict()\n","  labels = dict()\n","  for doc in np.unique(df['Bag']):\n","    doc_df = df[df['Bag'] == doc]\n","    doc_all_sentences = doc_df['Sentence'].values\n","    labels[doc] = doc_df['Class'].values[0]\n","    for i, sentence in enumerate(range(doc_all_sentences.shape[0])):\n","      if i == 0:\n","        doc_dict[doc] = list(doc_all_sentences[i])\n","      else:\n","        for n in list(doc_all_sentences[i]):\n","          doc_dict[doc].append(n)\n","    doc_dict[doc] = np.asarray(doc_dict[doc], dtype=np.int32)\n","\n","  df_new = pd.DataFrame(\n","      {\n","        'Bag': np.unique(df['Bag']),\n","        'Sentence': list(doc_dict.values()),\n","        'Class': list(labels.values())\n","      }\n","  )\n","\n","  return df_new\n","\n","\n","def zero_padding(df, max_words):\n","  labels = df['Class'].values\n","  \n","  # Pad sentences for 'max_words' words per sentence max.\n","  sentences = df['Sentence'].values\n","  X = pad_sequences(sentences, maxlen=max_words, dtype='float32', padding='post')\n","\n","  return X, labels"],"metadata":{"id":"0t52EE75sdgg"},"execution_count":null,"outputs":[]},{"cell_type":"markdown","source":["# MAIN"],"metadata":{"id":"R4XgUkxmTepz"}},{"cell_type":"code","source":["with zipfile.ZipFile(\"/content/Data.zip\", 'r') as zip_ref:\n","  zip_ref.extractall()"],"metadata":{"id":"5ZxSmztaV9GN"},"execution_count":null,"outputs":[]},{"cell_type":"markdown","source":["## 1. each document represented ***by the groups its sentences belong to***"],"metadata":{"id":"CvuW9CQtUuMX"}},{"cell_type":"code","source":["# Create bags of sentences for the train and test data.\n","train_df = create_bags_of_sentences(\"/content/Data/train-data.dat\",\n","                                    \"/content/Data/train-label.dat\")\n","test_df = create_bags_of_sentences(\"/content/Data/test-data.dat\",\n","                                   \"/content/Data/test-label.dat\")"],"metadata":{"id":"SvKd1cSuTfXN"},"execution_count":null,"outputs":[]},{"cell_type":"code","source":["train_df.head()"],"metadata":{"colab":{"base_uri":"https://localhost:8080/","height":206},"id":"0x2XdawoWaFI","executionInfo":{"status":"ok","timestamp":1684134754442,"user_tz":-180,"elapsed":8,"user":{"displayName":"Achil Pal","userId":"04608072963754583562"}},"outputId":"1d47f6b3-3f04-46d2-c344-e71c37d56fb7"},"execution_count":null,"outputs":[{"output_type":"execute_result","data":{"text/plain":["   Bag                                           Sentence  Class\n","0    0    [6705, 5997, 8310, 3606, 674, 8058, 5044, 4836]      1\n","1    0                           [4312, 5154, 8310, 4225]      1\n","2    1                            [1827, 1037, 8482, 483]      1\n","3    1  [3567, 6172, 6172, 2892, 1362, 787, 399, 777, ...      1\n","4    1      [318, 769, 4621, 3199, 1480, 6213, 971, 6890]      1"],"text/html":["\n","  <div id=\"df-788821c5-f1cd-4fc7-9315-0264c29657e1\">\n","    <div class=\"colab-df-container\">\n","      <div>\n","<style scoped>\n","    .dataframe tbody tr th:only-of-type {\n","        vertical-align: middle;\n","    }\n","\n","    .dataframe tbody tr th {\n","        vertical-align: top;\n","    }\n","\n","    .dataframe thead th {\n","        text-align: right;\n","    }\n","</style>\n","<table border=\"1\" class=\"dataframe\">\n","  <thead>\n","    <tr style=\"text-align: right;\">\n","      <th></th>\n","      <th>Bag</th>\n","      <th>Sentence</th>\n","      <th>Class</th>\n","    </tr>\n","  </thead>\n","  <tbody>\n","    <tr>\n","      <th>0</th>\n","      <td>0</td>\n","      <td>[6705, 5997, 8310, 3606, 674, 8058, 5044, 4836]</td>\n","      <td>1</td>\n","    </tr>\n","    <tr>\n","      <th>1</th>\n","      <td>0</td>\n","      <td>[4312, 5154, 8310, 4225]</td>\n","      <td>1</td>\n","    </tr>\n","    <tr>\n","      <th>2</th>\n","      <td>1</td>\n","      <td>[1827, 1037, 8482, 483]</td>\n","      <td>1</td>\n","    </tr>\n","    <tr>\n","      <th>3</th>\n","      <td>1</td>\n","      <td>[3567, 6172, 6172, 2892, 1362, 787, 399, 777, ...</td>\n","      <td>1</td>\n","    </tr>\n","    <tr>\n","      <th>4</th>\n","      <td>1</td>\n","      <td>[318, 769, 4621, 3199, 1480, 6213, 971, 6890]</td>\n","      <td>1</td>\n","    </tr>\n","  </tbody>\n","</table>\n","</div>\n","      <button class=\"colab-df-convert\" onclick=\"convertToInteractive('df-788821c5-f1cd-4fc7-9315-0264c29657e1')\"\n","              title=\"Convert this dataframe to an interactive table.\"\n","              style=\"display:none;\">\n","        \n","  <svg xmlns=\"http://www.w3.org/2000/svg\" height=\"24px\"viewBox=\"0 0 24 24\"\n","       width=\"24px\">\n","    <path d=\"M0 0h24v24H0V0z\" fill=\"none\"/>\n","    <path d=\"M18.56 5.44l.94 2.06.94-2.06 2.06-.94-2.06-.94-.94-2.06-.94 2.06-2.06.94zm-11 1L8.5 8.5l.94-2.06 2.06-.94-2.06-.94L8.5 2.5l-.94 2.06-2.06.94zm10 10l.94 2.06.94-2.06 2.06-.94-2.06-.94-.94-2.06-.94 2.06-2.06.94z\"/><path d=\"M17.41 7.96l-1.37-1.37c-.4-.4-.92-.59-1.43-.59-.52 0-1.04.2-1.43.59L10.3 9.45l-7.72 7.72c-.78.78-.78 2.05 0 2.83L4 21.41c.39.39.9.59 1.41.59.51 0 1.02-.2 1.41-.59l7.78-7.78 2.81-2.81c.8-.78.8-2.07 0-2.86zM5.41 20L4 18.59l7.72-7.72 1.47 1.35L5.41 20z\"/>\n","  </svg>\n","      </button>\n","      \n","  <style>\n","    .colab-df-container {\n","      display:flex;\n","      flex-wrap:wrap;\n","      gap: 12px;\n","    }\n","\n","    .colab-df-convert {\n","      background-color: #E8F0FE;\n","      border: none;\n","      border-radius: 50%;\n","      cursor: pointer;\n","      display: none;\n","      fill: #1967D2;\n","      height: 32px;\n","      padding: 0 0 0 0;\n","      width: 32px;\n","    }\n","\n","    .colab-df-convert:hover {\n","      background-color: #E2EBFA;\n","      box-shadow: 0px 1px 2px rgba(60, 64, 67, 0.3), 0px 1px 3px 1px rgba(60, 64, 67, 0.15);\n","      fill: #174EA6;\n","    }\n","\n","    [theme=dark] .colab-df-convert {\n","      background-color: #3B4455;\n","      fill: #D2E3FC;\n","    }\n","\n","    [theme=dark] .colab-df-convert:hover {\n","      background-color: #434B5C;\n","      box-shadow: 0px 1px 3px 1px rgba(0, 0, 0, 0.15);\n","      filter: drop-shadow(0px 1px 2px rgba(0, 0, 0, 0.3));\n","      fill: #FFFFFF;\n","    }\n","  </style>\n","\n","      <script>\n","        const buttonEl =\n","          document.querySelector('#df-788821c5-f1cd-4fc7-9315-0264c29657e1 button.colab-df-convert');\n","        buttonEl.style.display =\n","          google.colab.kernel.accessAllowed ? 'block' : 'none';\n","\n","        async function convertToInteractive(key) {\n","          const element = document.querySelector('#df-788821c5-f1cd-4fc7-9315-0264c29657e1');\n","          const dataTable =\n","            await google.colab.kernel.invokeFunction('convertToInteractive',\n","                                                     [key], {});\n","          if (!dataTable) return;\n","\n","          const docLinkHtml = 'Like what you see? Visit the ' +\n","            '<a target=\"_blank\" href=https://colab.research.google.com/notebooks/data_table.ipynb>data table notebook</a>'\n","            + ' to learn more about interactive tables.';\n","          element.innerHTML = '';\n","          dataTable['output_type'] = 'display_data';\n","          await google.colab.output.renderOutput(dataTable, element);\n","          const docLink = document.createElement('div');\n","          docLink.innerHTML = docLinkHtml;\n","          element.appendChild(docLink);\n","        }\n","      </script>\n","    </div>\n","  </div>\n","  "]},"metadata":{},"execution_count":7}]},{"cell_type":"code","source":["# Scale data\n","arr = []\n","for i in range(train_df.shape[0]):\n","  sentence = train_df[\"Sentence\"].values[i]\n","  for j in sentence:\n","    arr.append(j)\n","arr = np.array(arr)\n","\n","train_df[\"Sentence\"] = (train_df[\"Sentence\"] - arr.min()) / (arr.max() - arr.min())\n","test_df[\"Sentence\"] = (test_df[\"Sentence\"] - arr.min()) / (arr.max() - arr.min())"],"metadata":{"id":"yHS815Cqrqjw"},"execution_count":null,"outputs":[]},{"cell_type":"code","source":["train_df.head()"],"metadata":{"colab":{"base_uri":"https://localhost:8080/","height":206},"id":"WEbGpPmJsYYj","executionInfo":{"status":"ok","timestamp":1684134762711,"user_tz":-180,"elapsed":8,"user":{"displayName":"Achil Pal","userId":"04608072963754583562"}},"outputId":"d80b9d68-497f-47f4-cc42-6b039a8898df"},"execution_count":null,"outputs":[{"output_type":"execute_result","data":{"text/plain":["   Bag                                           Sentence  Class\n","0    0  [0.7870642094142505, 0.7039558633642446, 0.975...      1\n","1    0  [0.5061626951520132, 0.6050005869233478, 0.975...      1\n","2    1  [0.21446179129005752, 0.12172790233595493, 0.9...      1\n","3    1  [0.4187111163282075, 0.7244981805376218, 0.724...      1\n","4    1  [0.03732832492076535, 0.09026881089329733, 0.5...      1"],"text/html":["\n","  <div id=\"df-ea9f2042-6fd8-48db-b56e-26e548cc4e4c\">\n","    <div class=\"colab-df-container\">\n","      <div>\n","<style scoped>\n","    .dataframe tbody tr th:only-of-type {\n","        vertical-align: middle;\n","    }\n","\n","    .dataframe tbody tr th {\n","        vertical-align: top;\n","    }\n","\n","    .dataframe thead th {\n","        text-align: right;\n","    }\n","</style>\n","<table border=\"1\" class=\"dataframe\">\n","  <thead>\n","    <tr style=\"text-align: right;\">\n","      <th></th>\n","      <th>Bag</th>\n","      <th>Sentence</th>\n","      <th>Class</th>\n","    </tr>\n","  </thead>\n","  <tbody>\n","    <tr>\n","      <th>0</th>\n","      <td>0</td>\n","      <td>[0.7870642094142505, 0.7039558633642446, 0.975...</td>\n","      <td>1</td>\n","    </tr>\n","    <tr>\n","      <th>1</th>\n","      <td>0</td>\n","      <td>[0.5061626951520132, 0.6050005869233478, 0.975...</td>\n","      <td>1</td>\n","    </tr>\n","    <tr>\n","      <th>2</th>\n","      <td>1</td>\n","      <td>[0.21446179129005752, 0.12172790233595493, 0.9...</td>\n","      <td>1</td>\n","    </tr>\n","    <tr>\n","      <th>3</th>\n","      <td>1</td>\n","      <td>[0.4187111163282075, 0.7244981805376218, 0.724...</td>\n","      <td>1</td>\n","    </tr>\n","    <tr>\n","      <th>4</th>\n","      <td>1</td>\n","      <td>[0.03732832492076535, 0.09026881089329733, 0.5...</td>\n","      <td>1</td>\n","    </tr>\n","  </tbody>\n","</table>\n","</div>\n","      <button class=\"colab-df-convert\" onclick=\"convertToInteractive('df-ea9f2042-6fd8-48db-b56e-26e548cc4e4c')\"\n","              title=\"Convert this dataframe to an interactive table.\"\n","              style=\"display:none;\">\n","        \n","  <svg xmlns=\"http://www.w3.org/2000/svg\" height=\"24px\"viewBox=\"0 0 24 24\"\n","       width=\"24px\">\n","    <path d=\"M0 0h24v24H0V0z\" fill=\"none\"/>\n","    <path d=\"M18.56 5.44l.94 2.06.94-2.06 2.06-.94-2.06-.94-.94-2.06-.94 2.06-2.06.94zm-11 1L8.5 8.5l.94-2.06 2.06-.94-2.06-.94L8.5 2.5l-.94 2.06-2.06.94zm10 10l.94 2.06.94-2.06 2.06-.94-2.06-.94-.94-2.06-.94 2.06-2.06.94z\"/><path d=\"M17.41 7.96l-1.37-1.37c-.4-.4-.92-.59-1.43-.59-.52 0-1.04.2-1.43.59L10.3 9.45l-7.72 7.72c-.78.78-.78 2.05 0 2.83L4 21.41c.39.39.9.59 1.41.59.51 0 1.02-.2 1.41-.59l7.78-7.78 2.81-2.81c.8-.78.8-2.07 0-2.86zM5.41 20L4 18.59l7.72-7.72 1.47 1.35L5.41 20z\"/>\n","  </svg>\n","      </button>\n","      \n","  <style>\n","    .colab-df-container {\n","      display:flex;\n","      flex-wrap:wrap;\n","      gap: 12px;\n","    }\n","\n","    .colab-df-convert {\n","      background-color: #E8F0FE;\n","      border: none;\n","      border-radius: 50%;\n","      cursor: pointer;\n","      display: none;\n","      fill: #1967D2;\n","      height: 32px;\n","      padding: 0 0 0 0;\n","      width: 32px;\n","    }\n","\n","    .colab-df-convert:hover {\n","      background-color: #E2EBFA;\n","      box-shadow: 0px 1px 2px rgba(60, 64, 67, 0.3), 0px 1px 3px 1px rgba(60, 64, 67, 0.15);\n","      fill: #174EA6;\n","    }\n","\n","    [theme=dark] .colab-df-convert {\n","      background-color: #3B4455;\n","      fill: #D2E3FC;\n","    }\n","\n","    [theme=dark] .colab-df-convert:hover {\n","      background-color: #434B5C;\n","      box-shadow: 0px 1px 3px 1px rgba(0, 0, 0, 0.15);\n","      filter: drop-shadow(0px 1px 2px rgba(0, 0, 0, 0.3));\n","      fill: #FFFFFF;\n","    }\n","  </style>\n","\n","      <script>\n","        const buttonEl =\n","          document.querySelector('#df-ea9f2042-6fd8-48db-b56e-26e548cc4e4c button.colab-df-convert');\n","        buttonEl.style.display =\n","          google.colab.kernel.accessAllowed ? 'block' : 'none';\n","\n","        async function convertToInteractive(key) {\n","          const element = document.querySelector('#df-ea9f2042-6fd8-48db-b56e-26e548cc4e4c');\n","          const dataTable =\n","            await google.colab.kernel.invokeFunction('convertToInteractive',\n","                                                     [key], {});\n","          if (!dataTable) return;\n","\n","          const docLinkHtml = 'Like what you see? Visit the ' +\n","            '<a target=\"_blank\" href=https://colab.research.google.com/notebooks/data_table.ipynb>data table notebook</a>'\n","            + ' to learn more about interactive tables.';\n","          element.innerHTML = '';\n","          dataTable['output_type'] = 'display_data';\n","          await google.colab.output.renderOutput(dataTable, element);\n","          const docLink = document.createElement('div');\n","          docLink.innerHTML = docLinkHtml;\n","          element.appendChild(docLink);\n","        }\n","      </script>\n","    </div>\n","  </div>\n","  "]},"metadata":{},"execution_count":9}]},{"cell_type":"code","source":["# Get the bags and the labels.\n","max_words = 10\n","X_train_bag, y_train = create_bag_per_document(train_df, max_words)\n","X_test_bag, y_test = create_bag_per_document(test_df, max_words)"],"metadata":{"id":"x6vgmPIYWavZ"},"execution_count":null,"outputs":[]},{"cell_type":"code","source":["# a = combinations(enumerate(train_bag[:100]), 2)\n","# counter = 0\n","# for (i, x), (j, y) in a:\n","#   counter += 1\n","#   if counter == 1:\n","#     t0 = datetime.now()\n","#     a = max(directed_hausdorff(x, y)[0], directed_hausdorff(y, x)[0])\n","#     # print(hausdorff_symmetric(x, y))\n","#     t1 = datetime.now()\n","  \n","# print(counter * (t1 - t0))"],"metadata":{"id":"shJ_a3TLiwi-"},"execution_count":null,"outputs":[]},{"cell_type":"code","source":["N_train = -1\n","N_test = -1\n","\n","X_train_bag_sampled = X_train_bag[:N_train]\n","y_train_sampled = y_train[:N_train]\n","\n","X_test_bag_sampled = X_test_bag[:N_test]\n","y_test_sampled = y_test[:N_test]\n","\n","# Initialize distance matrix.\n","N_bags = X_train_bag_sampled.shape[0]\n","dist_matrix = np.zeros((N_bags, N_bags))\n","\n","# Calculate symmetric hausdorff distances.\n","for (i, x), (j, y) in combinations(enumerate(X_train_bag_sampled), 2):\n","  dist_matrix[i, j] = max(directed_hausdorff(x, y)[0], directed_hausdorff(y, x)[0])   # hausdorff_symmetric(x, y)\n","  dist_matrix[j, i] = dist_matrix[i, j]"],"metadata":{"id":"svwKzQljYWeQ"},"execution_count":null,"outputs":[]},{"cell_type":"code","source":["pd.DataFrame(dist_matrix).to_csv(\"dist_matrix_1.csv\")"],"metadata":{"id":"9-cblYvT9dl7"},"execution_count":null,"outputs":[]},{"cell_type":"code","source":["# K_list = range(3, 200, 40)\n","K_list = range(3, 20, 1)\n","mean_scores = medoids_grid_search_and_cv(dist_matrix, K_list, cv=4)\n","\n","plt.figure(figsize=(12, 4))\n","plt.plot(K_list, mean_scores,'--', color='r', marker='p')\n","plt.xlabel('K')\n","plt.ylabel('Silhouette Coefficient')\n","plt.title(f\"Medoids (Training Samples: {N_train})\")\n","plt.show()"],"metadata":{"id":"mfyuBc8efC_1","colab":{"base_uri":"https://localhost:8080/","height":447},"executionInfo":{"status":"ok","timestamp":1684147325331,"user_tz":-180,"elapsed":565639,"user":{"displayName":"Achil Pal","userId":"04608072963754583562"}},"outputId":"f4189303-bf4c-4b1b-a107-5260a5e6d755"},"execution_count":null,"outputs":[{"output_type":"stream","name":"stdout","text":["Best Silhouette score found after Grid search : 0.1859\n","Best parameter : 3\n"]},{"output_type":"display_data","data":{"text/plain":["<Figure size 1200x400 with 1 Axes>"],"image/png":"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\n"},"metadata":{}}]},{"cell_type":"code","source":["kmedoids = KMedoids(n_clusters=3, init='random')\n","kmedoids.fit(dist_matrix)\n","train_medoids = kmedoids.medoid_indices_\n","\n","X_train_transf = transform_bags(X_train_bag_sampled, train_medoids)\n","X_test_transf = transform_bags(X_train_bag_sampled, train_medoids, X_test_bag_sampled)\n","\n","print(X_train_transf.shape, X_test_transf.shape)"],"metadata":{"colab":{"base_uri":"https://localhost:8080/"},"id":"-4Y0raAiFitV","executionInfo":{"status":"ok","timestamp":1684147495653,"user_tz":-180,"elapsed":18380,"user":{"displayName":"Achil Pal","userId":"04608072963754583562"}},"outputId":"68aff866-64f6-473b-efa4-7b2928ac36ba"},"execution_count":null,"outputs":[{"output_type":"stream","name":"stdout","text":["(8250, 3) (3982, 3)\n"]}]},{"cell_type":"code","source":["# Visualize the new training data\n","fig = plt.figure()\n","ax = plt.axes(projection='3d')\n","\n","x = X_train_transf[:, 0]\n","y = X_train_transf[:, 1]\n","z = X_train_transf[:, 2]\n","\n","ax.scatter(x, y, z)\n","ax.set_xlabel('feature 1')\n","ax.set_ylabel('feature 2')\n","ax.set_zlabel('feature 3')\n","\n","plt.show()"],"metadata":{"colab":{"base_uri":"https://localhost:8080/","height":415},"id":"or9rQcnaI0QW","executionInfo":{"status":"ok","timestamp":1684147545852,"user_tz":-180,"elapsed":647,"user":{"displayName":"Achil Pal","userId":"04608072963754583562"}},"outputId":"ba149137-28a6-4c3c-c330-423960dc0b8b"},"execution_count":null,"outputs":[{"output_type":"display_data","data":{"text/plain":["<Figure size 640x480 with 1 Axes>"],"image/png":"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\n"},"metadata":{}}]},{"cell_type":"code","source":["names = ['LogisticRegression', 'MultinomialNB', 'RandomForest']\n","classifiers = [\n","    LogisticRegression(random_state=0, solver='lbfgs'),\n","    MultinomialNB(),\n","    RandomForestClassifier(n_estimators=30)\n","]\n","\n","for name, clf in zip(names, classifiers):\n","  print(f\"\\n\\n{name}\\n\")\n","  clf.fit(X_train_transf, y_train_sampled)\n","  y_preds = clf.predict(X_test_transf)\n","  print(classification_report(y_test_sampled, y_preds))"],"metadata":{"colab":{"base_uri":"https://localhost:8080/"},"id":"Sx5tL465UTh7","executionInfo":{"status":"ok","timestamp":1684147557294,"user_tz":-180,"elapsed":1713,"user":{"displayName":"Achil Pal","userId":"04608072963754583562"}},"outputId":"f2a57c61-e5a0-4eba-dca8-a5a236f91b45"},"execution_count":null,"outputs":[{"output_type":"stream","name":"stdout","text":["\n","\n","LogisticRegression\n","\n","              precision    recall  f1-score   support\n","\n","           0       0.61      1.00      0.76      2425\n","           1       0.00      0.00      0.00      1557\n","\n","    accuracy                           0.61      3982\n","   macro avg       0.30      0.50      0.38      3982\n","weighted avg       0.37      0.61      0.46      3982\n","\n","\n","\n","MultinomialNB\n","\n","              precision    recall  f1-score   support\n","\n","           0       0.61      1.00      0.76      2425\n","           1       0.00      0.00      0.00      1557\n","\n","    accuracy                           0.61      3982\n","   macro avg       0.30      0.50      0.38      3982\n","weighted avg       0.37      0.61      0.46      3982\n","\n","\n","\n","RandomForest\n","\n","              precision    recall  f1-score   support\n","\n","           0       0.62      0.79      0.69      2425\n","           1       0.41      0.23      0.29      1557\n","\n","    accuracy                           0.57      3982\n","   macro avg       0.51      0.51      0.49      3982\n","weighted avg       0.54      0.57      0.54      3982\n","\n"]}]},{"cell_type":"markdown","source":["Από τα παραπάνω αποτελέσματα παρατηρώ ότι:\n","* τα μοντέλα Logistic Regression και Multinomial Naive Bias πετυχαίνουν 61% accuracy, το οποίο είναι **λίγο υψηλότερο** από το αντίστοιχο του Random Forest (57%)\n","* τα μοντέλα Logistic Regression και Multinomial Naive Bias πετυχαίνουν 38% macro average, το οποίο είναι **αρκετά χαμηλότερο** από το αντίστοιχο του Random Forest (49%).\n","\n","Γενικά, τα παραπάνω ποσοστά είναι αρκετά χαμηλά. Ωστόσο, πρόκειται για ένα δύσκολο πρόβλημα και συνεπώς θεωρώ δεν είναι κάτι το οποίο θα έπρεπε να μας προβληματίσει."],"metadata":{"id":"v2Hvvz1RcL0O"}},{"cell_type":"markdown","source":["## 2. each document represented ***by all its sentences***"],"metadata":{"id":"1cb-MmYcgQuz"}},{"cell_type":"code","source":["# Create bags of sentences for the train and test data.\n","train_df = create_bags_of_sentences(\"/content/Data/train-data.dat\",\n","                                    \"/content/Data/train-label.dat\")\n","test_df = create_bags_of_sentences(\"/content/Data/test-data.dat\",\n","                                   \"/content/Data/test-label.dat\")"],"metadata":{"id":"VBhr8Vv4heYM"},"execution_count":null,"outputs":[]},{"cell_type":"code","source":["train_df.head()"],"metadata":{"colab":{"base_uri":"https://localhost:8080/","height":206},"executionInfo":{"status":"ok","timestamp":1684149316094,"user_tz":-180,"elapsed":4,"user":{"displayName":"Achil Pal","userId":"04608072963754583562"}},"outputId":"e4ca6a76-0c9a-471d-911b-eaec914dc90f","id":"9Egb838CheYM"},"execution_count":null,"outputs":[{"output_type":"execute_result","data":{"text/plain":["   Bag                                           Sentence  Class\n","0    0    [6705, 5997, 8310, 3606, 674, 8058, 5044, 4836]      1\n","1    0                           [4312, 5154, 8310, 4225]      1\n","2    1                            [1827, 1037, 8482, 483]      1\n","3    1  [3567, 6172, 6172, 2892, 1362, 787, 399, 777, ...      1\n","4    1      [318, 769, 4621, 3199, 1480, 6213, 971, 6890]      1"],"text/html":["\n","  <div id=\"df-4df100e4-813e-4ecd-a246-1a9c55876a9f\">\n","    <div class=\"colab-df-container\">\n","      <div>\n","<style scoped>\n","    .dataframe tbody tr th:only-of-type {\n","        vertical-align: middle;\n","    }\n","\n","    .dataframe tbody tr th {\n","        vertical-align: top;\n","    }\n","\n","    .dataframe thead th {\n","        text-align: right;\n","    }\n","</style>\n","<table border=\"1\" class=\"dataframe\">\n","  <thead>\n","    <tr style=\"text-align: right;\">\n","      <th></th>\n","      <th>Bag</th>\n","      <th>Sentence</th>\n","      <th>Class</th>\n","    </tr>\n","  </thead>\n","  <tbody>\n","    <tr>\n","      <th>0</th>\n","      <td>0</td>\n","      <td>[6705, 5997, 8310, 3606, 674, 8058, 5044, 4836]</td>\n","      <td>1</td>\n","    </tr>\n","    <tr>\n","      <th>1</th>\n","      <td>0</td>\n","      <td>[4312, 5154, 8310, 4225]</td>\n","      <td>1</td>\n","    </tr>\n","    <tr>\n","      <th>2</th>\n","      <td>1</td>\n","      <td>[1827, 1037, 8482, 483]</td>\n","      <td>1</td>\n","    </tr>\n","    <tr>\n","      <th>3</th>\n","      <td>1</td>\n","      <td>[3567, 6172, 6172, 2892, 1362, 787, 399, 777, ...</td>\n","      <td>1</td>\n","    </tr>\n","    <tr>\n","      <th>4</th>\n","      <td>1</td>\n","      <td>[318, 769, 4621, 3199, 1480, 6213, 971, 6890]</td>\n","      <td>1</td>\n","    </tr>\n","  </tbody>\n","</table>\n","</div>\n","      <button class=\"colab-df-convert\" onclick=\"convertToInteractive('df-4df100e4-813e-4ecd-a246-1a9c55876a9f')\"\n","              title=\"Convert this dataframe to an interactive table.\"\n","              style=\"display:none;\">\n","        \n","  <svg xmlns=\"http://www.w3.org/2000/svg\" height=\"24px\"viewBox=\"0 0 24 24\"\n","       width=\"24px\">\n","    <path d=\"M0 0h24v24H0V0z\" fill=\"none\"/>\n","    <path d=\"M18.56 5.44l.94 2.06.94-2.06 2.06-.94-2.06-.94-.94-2.06-.94 2.06-2.06.94zm-11 1L8.5 8.5l.94-2.06 2.06-.94-2.06-.94L8.5 2.5l-.94 2.06-2.06.94zm10 10l.94 2.06.94-2.06 2.06-.94-2.06-.94-.94-2.06-.94 2.06-2.06.94z\"/><path d=\"M17.41 7.96l-1.37-1.37c-.4-.4-.92-.59-1.43-.59-.52 0-1.04.2-1.43.59L10.3 9.45l-7.72 7.72c-.78.78-.78 2.05 0 2.83L4 21.41c.39.39.9.59 1.41.59.51 0 1.02-.2 1.41-.59l7.78-7.78 2.81-2.81c.8-.78.8-2.07 0-2.86zM5.41 20L4 18.59l7.72-7.72 1.47 1.35L5.41 20z\"/>\n","  </svg>\n","      </button>\n","      \n","  <style>\n","    .colab-df-container {\n","      display:flex;\n","      flex-wrap:wrap;\n","      gap: 12px;\n","    }\n","\n","    .colab-df-convert {\n","      background-color: #E8F0FE;\n","      border: none;\n","      border-radius: 50%;\n","      cursor: pointer;\n","      display: none;\n","      fill: #1967D2;\n","      height: 32px;\n","      padding: 0 0 0 0;\n","      width: 32px;\n","    }\n","\n","    .colab-df-convert:hover {\n","      background-color: #E2EBFA;\n","      box-shadow: 0px 1px 2px rgba(60, 64, 67, 0.3), 0px 1px 3px 1px rgba(60, 64, 67, 0.15);\n","      fill: #174EA6;\n","    }\n","\n","    [theme=dark] .colab-df-convert {\n","      background-color: #3B4455;\n","      fill: #D2E3FC;\n","    }\n","\n","    [theme=dark] .colab-df-convert:hover {\n","      background-color: #434B5C;\n","      box-shadow: 0px 1px 3px 1px rgba(0, 0, 0, 0.15);\n","      filter: drop-shadow(0px 1px 2px rgba(0, 0, 0, 0.3));\n","      fill: #FFFFFF;\n","    }\n","  </style>\n","\n","      <script>\n","        const buttonEl =\n","          document.querySelector('#df-4df100e4-813e-4ecd-a246-1a9c55876a9f button.colab-df-convert');\n","        buttonEl.style.display =\n","          google.colab.kernel.accessAllowed ? 'block' : 'none';\n","\n","        async function convertToInteractive(key) {\n","          const element = document.querySelector('#df-4df100e4-813e-4ecd-a246-1a9c55876a9f');\n","          const dataTable =\n","            await google.colab.kernel.invokeFunction('convertToInteractive',\n","                                                     [key], {});\n","          if (!dataTable) return;\n","\n","          const docLinkHtml = 'Like what you see? Visit the ' +\n","            '<a target=\"_blank\" href=https://colab.research.google.com/notebooks/data_table.ipynb>data table notebook</a>'\n","            + ' to learn more about interactive tables.';\n","          element.innerHTML = '';\n","          dataTable['output_type'] = 'display_data';\n","          await google.colab.output.renderOutput(dataTable, element);\n","          const docLink = document.createElement('div');\n","          docLink.innerHTML = docLinkHtml;\n","          element.appendChild(docLink);\n","        }\n","      </script>\n","    </div>\n","  </div>\n","  "]},"metadata":{},"execution_count":21}]},{"cell_type":"code","source":["train_df = create_one_bag_per_sentence(train_df)\n","test_df = create_one_bag_per_sentence(test_df)"],"metadata":{"id":"_VuQgjZMsKEa"},"execution_count":null,"outputs":[]},{"cell_type":"code","source":["train_df.head()"],"metadata":{"colab":{"base_uri":"https://localhost:8080/","height":206},"id":"RI63wsAZsW6Y","executionInfo":{"status":"ok","timestamp":1684149326831,"user_tz":-180,"elapsed":4,"user":{"displayName":"Achil Pal","userId":"04608072963754583562"}},"outputId":"622df205-49a9-4956-c592-b1250e50485a"},"execution_count":null,"outputs":[{"output_type":"execute_result","data":{"text/plain":["   Bag                                           Sentence  Class\n","0    0  [6705, 5997, 8310, 3606, 674, 8058, 5044, 4836...      1\n","1    1  [1827, 1037, 8482, 483, 3567, 6172, 6172, 2892...      1\n","2    2  [8482, 787, 2169, 4077, 6335, 1137, 1598, 1055...      0\n","3    3  [3843, 3843, 5427, 3355, 3843, 5154, 7451, 510...      0\n","4    4  [5877, 4688, 5877, 1137, 343, 4695, 1145, 125,...      0"],"text/html":["\n","  <div id=\"df-a7147536-5e22-432f-9c39-bac2c784dc6c\">\n","    <div class=\"colab-df-container\">\n","      <div>\n","<style scoped>\n","    .dataframe tbody tr th:only-of-type {\n","        vertical-align: middle;\n","    }\n","\n","    .dataframe tbody tr th {\n","        vertical-align: top;\n","    }\n","\n","    .dataframe thead th {\n","        text-align: right;\n","    }\n","</style>\n","<table border=\"1\" class=\"dataframe\">\n","  <thead>\n","    <tr style=\"text-align: right;\">\n","      <th></th>\n","      <th>Bag</th>\n","      <th>Sentence</th>\n","      <th>Class</th>\n","    </tr>\n","  </thead>\n","  <tbody>\n","    <tr>\n","      <th>0</th>\n","      <td>0</td>\n","      <td>[6705, 5997, 8310, 3606, 674, 8058, 5044, 4836...</td>\n","      <td>1</td>\n","    </tr>\n","    <tr>\n","      <th>1</th>\n","      <td>1</td>\n","      <td>[1827, 1037, 8482, 483, 3567, 6172, 6172, 2892...</td>\n","      <td>1</td>\n","    </tr>\n","    <tr>\n","      <th>2</th>\n","      <td>2</td>\n","      <td>[8482, 787, 2169, 4077, 6335, 1137, 1598, 1055...</td>\n","      <td>0</td>\n","    </tr>\n","    <tr>\n","      <th>3</th>\n","      <td>3</td>\n","      <td>[3843, 3843, 5427, 3355, 3843, 5154, 7451, 510...</td>\n","      <td>0</td>\n","    </tr>\n","    <tr>\n","      <th>4</th>\n","      <td>4</td>\n","      <td>[5877, 4688, 5877, 1137, 343, 4695, 1145, 125,...</td>\n","      <td>0</td>\n","    </tr>\n","  </tbody>\n","</table>\n","</div>\n","      <button class=\"colab-df-convert\" onclick=\"convertToInteractive('df-a7147536-5e22-432f-9c39-bac2c784dc6c')\"\n","              title=\"Convert this dataframe to an interactive table.\"\n","              style=\"display:none;\">\n","        \n","  <svg xmlns=\"http://www.w3.org/2000/svg\" height=\"24px\"viewBox=\"0 0 24 24\"\n","       width=\"24px\">\n","    <path d=\"M0 0h24v24H0V0z\" fill=\"none\"/>\n","    <path d=\"M18.56 5.44l.94 2.06.94-2.06 2.06-.94-2.06-.94-.94-2.06-.94 2.06-2.06.94zm-11 1L8.5 8.5l.94-2.06 2.06-.94-2.06-.94L8.5 2.5l-.94 2.06-2.06.94zm10 10l.94 2.06.94-2.06 2.06-.94-2.06-.94-.94-2.06-.94 2.06-2.06.94z\"/><path d=\"M17.41 7.96l-1.37-1.37c-.4-.4-.92-.59-1.43-.59-.52 0-1.04.2-1.43.59L10.3 9.45l-7.72 7.72c-.78.78-.78 2.05 0 2.83L4 21.41c.39.39.9.59 1.41.59.51 0 1.02-.2 1.41-.59l7.78-7.78 2.81-2.81c.8-.78.8-2.07 0-2.86zM5.41 20L4 18.59l7.72-7.72 1.47 1.35L5.41 20z\"/>\n","  </svg>\n","      </button>\n","      \n","  <style>\n","    .colab-df-container {\n","      display:flex;\n","      flex-wrap:wrap;\n","      gap: 12px;\n","    }\n","\n","    .colab-df-convert {\n","      background-color: #E8F0FE;\n","      border: none;\n","      border-radius: 50%;\n","      cursor: pointer;\n","      display: none;\n","      fill: #1967D2;\n","      height: 32px;\n","      padding: 0 0 0 0;\n","      width: 32px;\n","    }\n","\n","    .colab-df-convert:hover {\n","      background-color: #E2EBFA;\n","      box-shadow: 0px 1px 2px rgba(60, 64, 67, 0.3), 0px 1px 3px 1px rgba(60, 64, 67, 0.15);\n","      fill: #174EA6;\n","    }\n","\n","    [theme=dark] .colab-df-convert {\n","      background-color: #3B4455;\n","      fill: #D2E3FC;\n","    }\n","\n","    [theme=dark] .colab-df-convert:hover {\n","      background-color: #434B5C;\n","      box-shadow: 0px 1px 3px 1px rgba(0, 0, 0, 0.15);\n","      filter: drop-shadow(0px 1px 2px rgba(0, 0, 0, 0.3));\n","      fill: #FFFFFF;\n","    }\n","  </style>\n","\n","      <script>\n","        const buttonEl =\n","          document.querySelector('#df-a7147536-5e22-432f-9c39-bac2c784dc6c button.colab-df-convert');\n","        buttonEl.style.display =\n","          google.colab.kernel.accessAllowed ? 'block' : 'none';\n","\n","        async function convertToInteractive(key) {\n","          const element = document.querySelector('#df-a7147536-5e22-432f-9c39-bac2c784dc6c');\n","          const dataTable =\n","            await google.colab.kernel.invokeFunction('convertToInteractive',\n","                                                     [key], {});\n","          if (!dataTable) return;\n","\n","          const docLinkHtml = 'Like what you see? Visit the ' +\n","            '<a target=\"_blank\" href=https://colab.research.google.com/notebooks/data_table.ipynb>data table notebook</a>'\n","            + ' to learn more about interactive tables.';\n","          element.innerHTML = '';\n","          dataTable['output_type'] = 'display_data';\n","          await google.colab.output.renderOutput(dataTable, element);\n","          const docLink = document.createElement('div');\n","          docLink.innerHTML = docLinkHtml;\n","          element.appendChild(docLink);\n","        }\n","      </script>\n","    </div>\n","  </div>\n","  "]},"metadata":{},"execution_count":23}]},{"cell_type":"code","source":["lens = []\n","for i in range(train_df.shape[0]):\n","  lens.append(len(train_df[\"Sentence\"].values[i]))\n","\n","np.array(lens).max()"],"metadata":{"colab":{"base_uri":"https://localhost:8080/"},"id":"g-AVpQS-rPF4","executionInfo":{"status":"ok","timestamp":1684149332220,"user_tz":-180,"elapsed":439,"user":{"displayName":"Achil Pal","userId":"04608072963754583562"}},"outputId":"e0d3dfc6-9acf-48b0-e91c-e9491119df54"},"execution_count":null,"outputs":[{"output_type":"execute_result","data":{"text/plain":["270"]},"metadata":{},"execution_count":24}]},{"cell_type":"code","source":["# Scale data\n","arr = []\n","for i in range(train_df.shape[0]):\n","  sentence = train_df[\"Sentence\"].values[i]\n","  for j in sentence:\n","    arr.append(j)\n","arr = np.array(arr)\n","\n","train_df[\"Sentence\"]= (train_df[\"Sentence\"] - arr.min()) / (arr.max() - arr.min())\n","test_df[\"Sentence\"]= (test_df[\"Sentence\"] - arr.min()) / (arr.max() - arr.min())"],"metadata":{"id":"AMGU1PqCrPF5"},"execution_count":null,"outputs":[]},{"cell_type":"code","source":["train_df.head()"],"metadata":{"colab":{"base_uri":"https://localhost:8080/","height":206},"executionInfo":{"status":"ok","timestamp":1684149338758,"user_tz":-180,"elapsed":4,"user":{"displayName":"Achil Pal","userId":"04608072963754583562"}},"outputId":"34d64b73-f442-4a6e-d229-c22c613279d8","id":"HuTfqbY-rPF5"},"execution_count":null,"outputs":[{"output_type":"execute_result","data":{"text/plain":["   Bag                                           Sentence  Class\n","0    0  [0.7870642094142505, 0.7039558633642446, 0.975...      1\n","1    1  [0.21446179129005752, 0.12172790233595493, 0.9...      1\n","2    2  [0.9956567672262002, 0.09238173494541613, 0.25...      0\n","3    3  [0.45110928512736237, 0.45110928512736237, 0.6...      0\n","4    4  [0.689869703016786, 0.5502993309073835, 0.6898...      0"],"text/html":["\n","  <div id=\"df-a07e7e7f-95e5-4712-af37-13fc935cfff9\">\n","    <div class=\"colab-df-container\">\n","      <div>\n","<style scoped>\n","    .dataframe tbody tr th:only-of-type {\n","        vertical-align: middle;\n","    }\n","\n","    .dataframe tbody tr th {\n","        vertical-align: top;\n","    }\n","\n","    .dataframe thead th {\n","        text-align: right;\n","    }\n","</style>\n","<table border=\"1\" class=\"dataframe\">\n","  <thead>\n","    <tr style=\"text-align: right;\">\n","      <th></th>\n","      <th>Bag</th>\n","      <th>Sentence</th>\n","      <th>Class</th>\n","    </tr>\n","  </thead>\n","  <tbody>\n","    <tr>\n","      <th>0</th>\n","      <td>0</td>\n","      <td>[0.7870642094142505, 0.7039558633642446, 0.975...</td>\n","      <td>1</td>\n","    </tr>\n","    <tr>\n","      <th>1</th>\n","      <td>1</td>\n","      <td>[0.21446179129005752, 0.12172790233595493, 0.9...</td>\n","      <td>1</td>\n","    </tr>\n","    <tr>\n","      <th>2</th>\n","      <td>2</td>\n","      <td>[0.9956567672262002, 0.09238173494541613, 0.25...</td>\n","      <td>0</td>\n","    </tr>\n","    <tr>\n","      <th>3</th>\n","      <td>3</td>\n","      <td>[0.45110928512736237, 0.45110928512736237, 0.6...</td>\n","      <td>0</td>\n","    </tr>\n","    <tr>\n","      <th>4</th>\n","      <td>4</td>\n","      <td>[0.689869703016786, 0.5502993309073835, 0.6898...</td>\n","      <td>0</td>\n","    </tr>\n","  </tbody>\n","</table>\n","</div>\n","      <button class=\"colab-df-convert\" onclick=\"convertToInteractive('df-a07e7e7f-95e5-4712-af37-13fc935cfff9')\"\n","              title=\"Convert this dataframe to an interactive table.\"\n","              style=\"display:none;\">\n","        \n","  <svg xmlns=\"http://www.w3.org/2000/svg\" height=\"24px\"viewBox=\"0 0 24 24\"\n","       width=\"24px\">\n","    <path d=\"M0 0h24v24H0V0z\" fill=\"none\"/>\n","    <path d=\"M18.56 5.44l.94 2.06.94-2.06 2.06-.94-2.06-.94-.94-2.06-.94 2.06-2.06.94zm-11 1L8.5 8.5l.94-2.06 2.06-.94-2.06-.94L8.5 2.5l-.94 2.06-2.06.94zm10 10l.94 2.06.94-2.06 2.06-.94-2.06-.94-.94-2.06-.94 2.06-2.06.94z\"/><path d=\"M17.41 7.96l-1.37-1.37c-.4-.4-.92-.59-1.43-.59-.52 0-1.04.2-1.43.59L10.3 9.45l-7.72 7.72c-.78.78-.78 2.05 0 2.83L4 21.41c.39.39.9.59 1.41.59.51 0 1.02-.2 1.41-.59l7.78-7.78 2.81-2.81c.8-.78.8-2.07 0-2.86zM5.41 20L4 18.59l7.72-7.72 1.47 1.35L5.41 20z\"/>\n","  </svg>\n","      </button>\n","      \n","  <style>\n","    .colab-df-container {\n","      display:flex;\n","      flex-wrap:wrap;\n","      gap: 12px;\n","    }\n","\n","    .colab-df-convert {\n","      background-color: #E8F0FE;\n","      border: none;\n","      border-radius: 50%;\n","      cursor: pointer;\n","      display: none;\n","      fill: #1967D2;\n","      height: 32px;\n","      padding: 0 0 0 0;\n","      width: 32px;\n","    }\n","\n","    .colab-df-convert:hover {\n","      background-color: #E2EBFA;\n","      box-shadow: 0px 1px 2px rgba(60, 64, 67, 0.3), 0px 1px 3px 1px rgba(60, 64, 67, 0.15);\n","      fill: #174EA6;\n","    }\n","\n","    [theme=dark] .colab-df-convert {\n","      background-color: #3B4455;\n","      fill: #D2E3FC;\n","    }\n","\n","    [theme=dark] .colab-df-convert:hover {\n","      background-color: #434B5C;\n","      box-shadow: 0px 1px 3px 1px rgba(0, 0, 0, 0.15);\n","      filter: drop-shadow(0px 1px 2px rgba(0, 0, 0, 0.3));\n","      fill: #FFFFFF;\n","    }\n","  </style>\n","\n","      <script>\n","        const buttonEl =\n","          document.querySelector('#df-a07e7e7f-95e5-4712-af37-13fc935cfff9 button.colab-df-convert');\n","        buttonEl.style.display =\n","          google.colab.kernel.accessAllowed ? 'block' : 'none';\n","\n","        async function convertToInteractive(key) {\n","          const element = document.querySelector('#df-a07e7e7f-95e5-4712-af37-13fc935cfff9');\n","          const dataTable =\n","            await google.colab.kernel.invokeFunction('convertToInteractive',\n","                                                     [key], {});\n","          if (!dataTable) return;\n","\n","          const docLinkHtml = 'Like what you see? Visit the ' +\n","            '<a target=\"_blank\" href=https://colab.research.google.com/notebooks/data_table.ipynb>data table notebook</a>'\n","            + ' to learn more about interactive tables.';\n","          element.innerHTML = '';\n","          dataTable['output_type'] = 'display_data';\n","          await google.colab.output.renderOutput(dataTable, element);\n","          const docLink = document.createElement('div');\n","          docLink.innerHTML = docLinkHtml;\n","          element.appendChild(docLink);\n","        }\n","      </script>\n","    </div>\n","  </div>\n","  "]},"metadata":{},"execution_count":26}]},{"cell_type":"code","source":["# Get the bags and the labels.\n","max_words = 270\n","X_train_bag, y_train = zero_padding(train_df, max_words)\n","X_test_bag, y_test = zero_padding(test_df, max_words)"],"metadata":{"id":"qMyTcJPct_jH"},"execution_count":null,"outputs":[]},{"cell_type":"code","source":["X_train_bag_sampled.shape"],"metadata":{"colab":{"base_uri":"https://localhost:8080/"},"id":"-aZsaea0ymD5","executionInfo":{"status":"ok","timestamp":1684149347184,"user_tz":-180,"elapsed":3,"user":{"displayName":"Achil Pal","userId":"04608072963754583562"}},"outputId":"7cf11fa5-ab07-406e-eae4-f005087cd557"},"execution_count":null,"outputs":[{"output_type":"execute_result","data":{"text/plain":["(8250,)"]},"metadata":{},"execution_count":28}]},{"cell_type":"code","source":["N_train = -1\n","N_test = -1\n","\n","X_train_bag_sampled = X_train_bag[:N_train]\n","y_train_sampled = y_train[:N_train]\n","\n","X_test_bag_sampled = X_test_bag[:N_test]\n","y_test_sampled = y_test[:N_test]\n","\n","# Initialize distance matrix.\n","N_bags = X_train_bag_sampled.shape[0]\n","dist_matrix = np.zeros((N_bags, N_bags))\n","\n","for i in range(N_bags):\n","  x = X_train_bag_sampled[i].reshape(1, -1)\n","  for j in range(N_bags):\n","    y = X_train_bag_sampled[j].reshape(1, -1)\n","    dist_matrix[i, j] = max(directed_hausdorff(x, y)[0], directed_hausdorff(y, x)[0])   # hausdorff_symmetric(x, y)\n","\n","# # Calculate symmetric hausdorff distances.\n","# for (i, x), (j, y) in combinations(enumerate(X_train_bag_sampled), 2):\n","#   dist_matrix[i, j] = max(directed_hausdorff(x, y)[0], directed_hausdorff(y, x)[0])   # hausdorff_symmetric(x, y)\n","#   dist_matrix[j, i] = dist_matrix[i, j]"],"metadata":{"id":"mTSLhLu8yQ_N"},"execution_count":null,"outputs":[]},{"cell_type":"code","source":["# K_list = range(3, 200, 40)\n","K_list = range(2, 20, 1)\n","mean_scores = medoids_grid_search_and_cv(dist_matrix, K_list, cv=4)\n","\n","plt.figure(figsize=(12, 4))\n","plt.plot(K_list, mean_scores,'--', color='r', marker='p')\n","plt.xlabel('K')\n","plt.ylabel('Silhouette Coefficient')\n","plt.title(f\"Medoids (Training Samples: {N_train})\")\n","plt.show()"],"metadata":{"colab":{"base_uri":"https://localhost:8080/"},"executionInfo":{"status":"ok","timestamp":1684173090843,"user_tz":-180,"elapsed":844031,"user":{"displayName":"Achil Pal","userId":"04608072963754583562"}},"outputId":"94def665-2797-4372-9413-55126c9c371f","id":"xbD9XJjEyQ_O"},"execution_count":null,"outputs":[{"output_type":"stream","name":"stdout","text":["Best Silhouette score found after Grid search : 0.6680\n","Best parameter : 2\n"]},{"output_type":"display_data","data":{"text/plain":["<Figure size 1200x400 with 1 Axes>"],"image/png":"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\n"},"metadata":{}}]},{"cell_type":"code","source":["kmedoids = KMedoids(n_clusters=3, init='random')\n","kmedoids.fit(dist_matrix)\n","train_medoids = kmedoids.medoid_indices_\n","\n","X_train_transf = transform_bags(X_train_bag_sampled, train_medoids)\n","X_test_transf = transform_bags(X_train_bag_sampled, train_medoids, X_test_bag_sampled)\n","\n","print(X_train_transf.shape, X_test_transf.shape)"],"metadata":{"colab":{"base_uri":"https://localhost:8080/"},"executionInfo":{"status":"ok","timestamp":1684173109341,"user_tz":-180,"elapsed":18502,"user":{"displayName":"Achil Pal","userId":"04608072963754583562"}},"outputId":"6ecc49d5-5ec4-47f4-83b6-e8cd4f5ee005","id":"HAhv4ght1AOG"},"execution_count":null,"outputs":[{"output_type":"stream","name":"stdout","text":["(8250, 3) (3982, 3)\n"]}]},{"cell_type":"code","source":["# Visualize the new training data\n","fig = plt.figure()\n","ax = plt.axes(projection='3d')\n","\n","x = X_train_transf[:, 0]\n","y = X_train_transf[:, 1]\n","z = X_train_transf[:, 2]\n","\n","ax.scatter(x, y, z)\n","ax.set_xlabel('feature 1')\n","ax.set_ylabel('feature 2')\n","ax.set_zlabel('feature 3')\n","\n","plt.show()"],"metadata":{"colab":{"base_uri":"https://localhost:8080/"},"executionInfo":{"status":"ok","timestamp":1684173109341,"user_tz":-180,"elapsed":13,"user":{"displayName":"Achil Pal","userId":"04608072963754583562"}},"outputId":"f964ce4e-d384-464e-dd6d-cb1ee4f89931","id":"IkkQvNZX1AOG"},"execution_count":null,"outputs":[{"output_type":"display_data","data":{"text/plain":["<Figure size 640x480 with 1 Axes>"],"image/png":"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\n"},"metadata":{}}]},{"cell_type":"code","source":["names = ['LogisticRegression', 'MultinomialNB', 'RandomForest']\n","classifiers = [\n","    LogisticRegression(random_state=0, solver='lbfgs'),\n","    MultinomialNB(),\n","    RandomForestClassifier(n_estimators=30)\n","]\n","\n","for name, clf in zip(names, classifiers):\n","  print(f\"\\n\\n{name}\\n\")\n","  clf.fit(X_train_transf, y_train_sampled)\n","  y_preds = clf.predict(X_test_transf)\n","  print(classification_report(y_test_sampled, y_preds))"],"metadata":{"colab":{"base_uri":"https://localhost:8080/"},"executionInfo":{"status":"ok","timestamp":1684173110519,"user_tz":-180,"elapsed":1189,"user":{"displayName":"Achil Pal","userId":"04608072963754583562"}},"outputId":"195e6cbc-bd26-497d-80d3-688740177341","id":"rWdei6IW1AOG"},"execution_count":null,"outputs":[{"output_type":"stream","name":"stdout","text":["\n","\n","LogisticRegression\n","\n","              precision    recall  f1-score   support\n","\n","           0       0.61      1.00      0.76      2425\n","           1       0.00      0.00      0.00      1557\n","\n","    accuracy                           0.61      3982\n","   macro avg       0.30      0.50      0.38      3982\n","weighted avg       0.37      0.61      0.46      3982\n","\n","\n","\n","MultinomialNB\n","\n","              precision    recall  f1-score   support\n","\n","           0       0.61      1.00      0.76      2425\n","           1       0.00      0.00      0.00      1557\n","\n","    accuracy                           0.61      3982\n","   macro avg       0.30      0.50      0.38      3982\n","weighted avg       0.37      0.61      0.46      3982\n","\n","\n","\n","RandomForest\n","\n","              precision    recall  f1-score   support\n","\n","           0       0.61      0.77      0.68      2425\n","           1       0.40      0.24      0.30      1557\n","\n","    accuracy                           0.56      3982\n","   macro avg       0.51      0.51      0.49      3982\n","weighted avg       0.53      0.56      0.53      3982\n","\n"]}]},{"cell_type":"markdown","source":["Από τα παραπάνω αποτελέσματα παρατηρώ ότι:\n","* τα μοντέλα Logistic Regression και Multinomial Naive Bias πετυχαίνουν 61% accuracy, το οποίο είναι **λίγο υψηλότερο** από το αντίστοιχο του Random Forest (56%)\n","* τα μοντέλα Logistic Regression και Multinomial Naive Bias πετυχαίνουν 38% macro average, το οποίο είναι **αρκετά χαμηλότερο** από το αντίστοιχο του Random Forest (49%).\n","\n","Γενικά, τα παραπάνω ποσοστά είναι αρκετά χαμηλά. Ωστόσο, πρόκειται για ένα δύσκολο πρόβλημα και συνεπώς θεωρώ δεν είναι κάτι το οποίο θα έπρεπε να μας προβληματίσει."],"metadata":{"id":"QO-4zOx9d4o6"}},{"cell_type":"markdown","source":["***Συνοψίζοντας, συγκρίνοντας τα αποτελέσματα των 2 παραπάνω μεθόδων, θα μπορούσαμε να πούμε ότι δεν υπάρχουν ουσιαστικές διαφορές. Συνεπώς, για το συγκεκριμένο πρόβλημα, δεν είναι κάποια μέθοδος προτιμότερη από την άλλη.***"],"metadata":{"id":"a1aMsT3UeWbh"}}]}