experiment_generalization_ability.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "id": "dfe1dd5c-8092-4e97-b7ca-31080ee8b588",
   "metadata": {},
   "outputs": [],
   "source": [
    "import numpy as np\n",
    "import pandas as pd\n",
    "from matplotlib import pyplot as plt\n",
    "import seaborn as sns"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "id": "70176221-f818-45a1-bb95-faae15a46336",
   "metadata": {},
   "outputs": [],
   "source": [
    "data_lst = []\n",
    "y_lst = []\n",
    "for type in ['point', 'pattern']:\n",
    "    df = pd.read_csv(f'data_processed/synthetic_{type}.csv', index_col=0)\n",
    "    split = int(0.4*df.shape[0])\n",
    "    train_df = df.iloc[: split]\n",
    "    test_df = df.iloc[split: ]\n",
    "    label01 = df['label01'].values[split:]\n",
    "    labelmul = df['labelmul'].values[split:]\n",
    "    df = df.drop(['label01', 'labelmul'], axis=1)\n",
    "\n",
    "    y = label01\n",
    "    data = test_df.values[:, 0]\n",
    "\n",
    "    if type == 'pattern':\n",
    "        anom_pairs = []\n",
    "        if y is not None:\n",
    "            anom_index = np.where(y == 1)[0]\n",
    "            tmp_seg = []\n",
    "            for i in anom_index:\n",
    "                tmp_seg.append(i)\n",
    "                if i + 1 not in anom_index:\n",
    "                    anom_pairs.append((tmp_seg[0], tmp_seg[-1]))\n",
    "                    tmp_seg = []\n",
    "    \n",
    "    data_lst.append(data)\n",
    "    y_lst.append(y)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "id": "df5fbc46-cd69-47d0-b592-d34603738feb",
   "metadata": {},
   "outputs": [],
   "source": [
    "path = '@results_showcase/'\n",
    "algo_lst = ['couta']\n",
    "algo_lst = algo_lst[::-1]\n",
    "n_algo = len(algo_lst)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "id": "e9857a25-a6cc-465f-b70d-348dbdc9d7dd",
   "metadata": {},
   "outputs": [],
   "source": [
    "\n",
    "left = 50\n",
    "right = 500\n",
    "\n",
    "\n",
    "score_df_lst = []\n",
    "for type in ['point', 'pattern']:\n",
    "    score_df = pd.DataFrame()\n",
    "    for algo in algo_lst:\n",
    "        # print(algo)\n",
    "        try:\n",
    "            score = np.load(path + f'showcase_{type}_score_{algo}.npy', allow_pickle=True)\n",
    "            score_ = score[left:right]\n",
    "            score = (score - np.min(score_)) / (np.max(score_) - np.min(score_))\n",
    "        except FileNotFoundError:\n",
    "            score = np.zeros(len(y))\n",
    "            print(type, algo, 'missing')\n",
    "        score_df[algo] = score\n",
    "    index = np.arange(len(score))\n",
    "    score_df['index'] = index\n",
    "    \n",
    "    score_df['COUTA (ours)'] = score_df['couta']\n",
    "    score_df = score_df.drop('couta', axis=1)\n",
    "    \n",
    "    score_df_lst.append(score_df)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "id": "bb367844-ae3a-4c01-946f-6747367f74b1",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>index</th>\n",
       "      <th>COUTA (ours)</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>0</td>\n",
       "      <td>-0.000128</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>1</td>\n",
       "      <td>-0.000128</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>2</td>\n",
       "      <td>-0.000128</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>3</td>\n",
       "      <td>-0.000128</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>4</td>\n",
       "      <td>-0.000128</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>595</th>\n",
       "      <td>595</td>\n",
       "      <td>0.000136</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>596</th>\n",
       "      <td>596</td>\n",
       "      <td>0.000396</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>597</th>\n",
       "      <td>597</td>\n",
       "      <td>0.000464</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>598</th>\n",
       "      <td>598</td>\n",
       "      <td>0.000173</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>599</th>\n",
       "      <td>599</td>\n",
       "      <td>0.000854</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>600 rows × 2 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "     index  COUTA (ours)\n",
       "0        0     -0.000128\n",
       "1        1     -0.000128\n",
       "2        2     -0.000128\n",
       "3        3     -0.000128\n",
       "4        4     -0.000128\n",
       "..     ...           ...\n",
       "595    595      0.000136\n",
       "596    596      0.000396\n",
       "597    597      0.000464\n",
       "598    598      0.000173\n",
       "599    599      0.000854\n",
       "\n",
       "[600 rows x 2 columns]"
      ]
     },
     "execution_count": 14,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "score_df"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "id": "d1de276b-73c8-4dcd-9b31-9c0f2f92cadb",
   "metadata": {},
   "outputs": [],
   "source": [
    "algo_lst = ['COUTA (ours)']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "id": "39a86023-8a6e-47ff-befa-7af657a0c293",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 864x129.6 with 4 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "fig = plt.figure(figsize=(12, 1.8*n_algo))\n",
    "\n",
    "for ii, (y, data, score_df) in enumerate(zip(y_lst, data_lst, score_df_lst)):\n",
    "    fig.add_subplot(n_algo+1, 2, 1+ii)\n",
    "    index = score_df['index']\n",
    "    sns.lineplot(x=index, y=data, linewidth=0.8, color=sns.color_palette('Greys_r')[0])\n",
    "    \n",
    "    if ii == 0:\n",
    "        plt.scatter(np.where(y==1)[0], data[np.where(y==1)[0]], color='none', marker='o', linewidth=0.8,\n",
    "                    alpha=1, edgecolors='r', s=40)\n",
    "        for idx in np.where(y==1)[0]:\n",
    "            plt.axvline(idx, color='darkgoldenrod', alpha=1, linewidth=0.5,  linestyle='--')\n",
    "    else:\n",
    "        value_min = np.around(data.min(), decimals=1) - 0.1;value_max = np.around(data.max(), decimals=1) + 0.1\n",
    "        values_ = np.linspace(value_min, value_max, 10)\n",
    "        for pair in anom_pairs:\n",
    "            x1 = np.ones(len(values_)) * pair[0];x2 = np.ones(len(values_)) * pair[1]\n",
    "            plt.fill_betweenx(values_, x1, x2, alpha=0.3, color='goldenrod')\n",
    "\n",
    "    ax = plt.gca()\n",
    "    for loc in ['bottom', 'left', 'top', 'right']:\n",
    "        ax.spines[loc].set_linewidth(0.6)\n",
    "\n",
    "     \n",
    "    plt.xlim(left, right)\n",
    "    plt.xlabel('')\n",
    "    plt.ylabel('Data' if ii == 0 else '', fontsize=16)\n",
    "    plt.xticks([])\n",
    "    plt.yticks([])\n",
    "    plt.title('Time-series Data with Point-wise \\n Anomalies (Global/Contextual)' if ii ==0 else 'Time-series Data with Pattern-wise \\n Anomalies (Seasonal/Shapelet)', fontsize=18)\n",
    "\n",
    "    for jj, algo in enumerate(algo_lst):\n",
    "        \n",
    "        ax=fig.add_subplot(n_algo+1, 2, 3+ii+jj*2)\n",
    "        # palette = sns.color_palette(\"bone\", 6)\n",
    "\n",
    "        score_df['name'] = algo\n",
    "        sns.lineplot(x='index', y=algo, data=score_df, hue='name', legend=True, \n",
    "                     linewidth=0.8, \n",
    "                     # palette=sns.color_palette(\"ocean\", 1)\n",
    "                     palette=sns.color_palette(\"cividis\", 10)[:1]\n",
    "                    )\n",
    "        plt.legend(loc='upper left', ncol=1, fontsize=16, handlelength=0.3)\n",
    "        if ii == 0:   \n",
    "            for idx in np.where(y==1)[0]:\n",
    "                plt.axvline(idx, color='darkgoldenrod', alpha=1, linewidth=0.5, linestyle='--')\n",
    "        else:\n",
    "            values_ = np.linspace(-0.02, 1.1, 10)\n",
    "            for pair in anom_pairs:\n",
    "                plt.fill_betweenx(values_, np.ones(len(values_))*pair[0], np.ones(len(values_))*pair[1], alpha=0.3, \n",
    "                                  # color='red'\n",
    "                                  color='goldenrod'\n",
    "                                 )\n",
    "\n",
    "        plt.xlim(left, right)\n",
    "        plt.ylim(-0.02, 1.1)\n",
    "        plt.xlabel('')\n",
    "        plt.ylabel('Scores' if ii == 0 else '', fontsize=16)\n",
    "        plt.xticks([])\n",
    "        plt.yticks([])\n",
    "        \n",
    "        ax = plt.gca()\n",
    "        for loc in ['bottom', 'left', 'top', 'right']:\n",
    "            ax.spines[loc].set_linewidth(0.6)\n",
    "\n",
    "\n",
    "plt.subplots_adjust(wspace=0.1, hspace=0.12)\n",
    "# plt.savefig(f'showcase.pdf', bbox_inches = 'tight')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "3e165898-8012-444a-8647-fa2a971d3d05",
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "cuda113",
   "language": "python",
   "name": "cuda113"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.8.0"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}