DRACO.ipynb

{
  "cells": [
    {
      "cell_type": "code",
      "execution_count": 2,
      "metadata": {
        "id": "0FJcJ5v1bVWN"
      },
      "outputs": [],
      "source": [
        "import numpy as np\n",
        "import matplotlib.pyplot as plt\n",
        "import os\n",
        "import glob\n",
        "from pyntcloud import PyntCloud\n",
        "import plyfile as ply\n",
        "import sys\n",
        "import DracoPy\n",
        "import shutil\n",
        "import pandas as pd\n",
        "import open3d as o3d\n",
        "from tqdm import tqdm"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "Thojbe9dcjdz"
      },
      "source": [
        "Draco compression over the whole merged directory"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 3,
      "metadata": {},
      "outputs": [],
      "source": [
        "import time\n",
        "import plyfile\n",
        "\n",
        "path_to_combined_data = 'dataset-downloader-kit/CV/dataset/Town01_Opt_ClearSunset/dataset_final'\n",
        "all_pc = glob.glob(os.path.join(path_to_combined_data, '*.ply'))\n"
      ]
    },
    {
      "attachments": {},
      "cell_type": "markdown",
      "metadata": {},
      "source": [
        "## Quality metric\n",
        "\n",
        "It is possible to define different quality metrics to evaluate the quality of the compression in a pointcloud scenario.\n",
        "\n",
        "In this case we decide to apply, mostly for computational reasons, the mean point distance error.\n"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 4,
      "metadata": {},
      "outputs": [],
      "source": [
        "from scipy.spatial import KDTree\n",
        "def mean_point_distance_error(points, decoded_points):\n",
        "\n",
        "    # Create a KDTree of the decoded points\n",
        "    decoded_points_tree = KDTree(decoded_points)\n",
        "    # Find the closest point in the decoded point cloud for each point in the original point cloud\n",
        "    distances, closest_indices = decoded_points_tree.query(points)\n",
        "    # Get the closest points\n",
        "    closest_points = decoded_points[closest_indices]\n",
        "    # Calculate the mean point distance error\n",
        "    mpde = np.mean(np.linalg.norm(points - closest_points, axis=1))\n",
        "    return mpde"
      ]
    },
    {
      "attachments": {},
      "cell_type": "markdown",
      "metadata": {},
      "source": [
        "# Unlabelled pointcloud compression\n",
        "\n",
        "Let us first compress the unlabelled pointclouds. We compress each of the 100 pointclouds obtained by merginf top, left and right LIDARS and compute the following metric:\n",
        "\n",
        "- Compression ratio\n",
        "- Mean point distance error\n",
        "- Number of lost points\n",
        "- Computation time\n",
        "\n",
        "Note that compression ratio is computed as the ratio between the size of the compressed file and the size of the original file:\n",
        "\n",
        "$$ \\frac{size_{compressed}}{size_{original}} $$\n",
        "\n",
        "Therefore, it represents how much the file has been compressed. The lower the better."
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "hG3OTgxWciHS",
        "outputId": "5f476f5b-30e8-4f95-9b68-f91ea3a76e22"
      },
      "outputs": [],
      "source": [
        "draco_unlabeled = pd.DataFrame(columns=['filename', 'original_size', 'original_size_bytes', \n",
        "                                        'encoded_size', \n",
        "                                        'compression_ratio', \n",
        "                                        'encoding_decoding_time', 'lost_points', 'mpde'])\n",
        "\n",
        "for pc in all_pc:\n",
        "\n",
        "    # Start the timer\n",
        "    start = time.time()\n",
        "\n",
        "    # Read the ply file\n",
        "    plydata = plyfile.PlyData.read(pc)\n",
        "    # Extract the XYZ coordinates of the points\n",
        "    points = np.vstack([plydata['vertex']['x'], plydata['vertex']['y'], plydata['vertex']['z']]).T\n",
        "\n",
        "    # Measure the number of points in the original point cloud\n",
        "    original_size = len(points)\n",
        "    original_size_bytes = len(points.tobytes())\n",
        "\n",
        "\n",
        "    # Encode the point cloud\n",
        "    encoded_data = DracoPy.encode(points)\n",
        "    encoded_size_bytes = len(encoded_data)\n",
        "\n",
        "    # Measure the size of the encoded data\n",
        "    encoded_size = len(encoded_data)\n",
        "     # Calculate the compression ratio\n",
        "    encoded_compression_ratio = (encoded_size_bytes / original_size_bytes)\n",
        "    \n",
        "    print (f\"Original size: {original_size_bytes} points\\tEncoded size: {encoded_size_bytes} bytes, \\t Compression ratio: {encoded_compression_ratio}\")\n",
        "\n",
        "    # Decode the encoded data\n",
        "    decoded_points = DracoPy.decode(encoded_data)\n",
        "    # End the timer\n",
        "    end = time.time()\n",
        "\n",
        "    # save the decoded points\n",
        "    decoded_points_filename = os.path.join('results', 'draco_unlabeled_decoded')\n",
        "\n",
        "    # save the decoded points \n",
        "    if not os.path.exists(decoded_points_filename):\n",
        "        os.makedirs(decoded_points_filename)\n",
        "    \n",
        "    decoded_points_filename = os.path.join(decoded_points_filename, os.path.basename(pc))\n",
        "    ply_decoded = plyfile.PlyData(\\\n",
        "        [plyfile.PlyElement.describe(\\\n",
        "            np.array(list(zip(decoded_points.points[:,0], \n",
        "            decoded_points.points[:,1], \n",
        "            decoded_points.points[:,2])), \n",
        "            dtype=[('x', 'f4'), ('y', 'f4'), ('z', 'f4')]), 'vertex')])\n",
        "            \n",
        "    ply_decoded.write(decoded_points_filename + '.ply')\n",
        "\n",
        "    #metric for quality of the decoded points\n",
        "    mpde = mean_point_distance_error(points, decoded_points.points)\n",
        "\n",
        "    # fiond nearest points to the decoded points in the original point cloud\n",
        "    lost_points = len(points) - len(decoded_points.points)\n",
        "\n",
        "    # Save the results\n",
        "    draco_unlabeled = draco_unlabeled.append({'filename': pc, \n",
        "                                            'original_size': original_size, \n",
        "                                            'original_size_bytes': original_size_bytes, \n",
        "                                            'encoded_size': encoded_size, \n",
        "                                            'compression_ratio': encoded_compression_ratio, \n",
        "                                            'encoding_decoding_time': end - start, \n",
        "                                            'lost_points': lost_points,\n",
        "                                            'mpde': mpde\n",
        "                                            }, \n",
        "                                            ignore_index=True)\n",
        "\n",
        "# save the results\n",
        "draco_unlabeled.to_csv('results/draco_unlabeled.csv', index=False)\n"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 3,
      "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>filename</th>\n",
              "      <th>original_size_bytes</th>\n",
              "      <th>encoded_size</th>\n",
              "      <th>compression_ratio</th>\n",
              "      <th>encoding_decoding_time</th>\n",
              "      <th>lost_points</th>\n",
              "      <th>mpde</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>dataset-downloader-kit/CV/dataset/Town01_Opt_C...</td>\n",
              "      <td>3416088</td>\n",
              "      <td>314512</td>\n",
              "      <td>0.092068</td>\n",
              "      <td>0.615055</td>\n",
              "      <td>0</td>\n",
              "      <td>0.005322</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1</th>\n",
              "      <td>dataset-downloader-kit/CV/dataset/Town01_Opt_C...</td>\n",
              "      <td>3251676</td>\n",
              "      <td>299772</td>\n",
              "      <td>0.092190</td>\n",
              "      <td>0.452345</td>\n",
              "      <td>0</td>\n",
              "      <td>0.005561</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2</th>\n",
              "      <td>dataset-downloader-kit/CV/dataset/Town01_Opt_C...</td>\n",
              "      <td>3462324</td>\n",
              "      <td>334056</td>\n",
              "      <td>0.096483</td>\n",
              "      <td>0.349651</td>\n",
              "      <td>1316</td>\n",
              "      <td>0.006069</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3</th>\n",
              "      <td>dataset-downloader-kit/CV/dataset/Town01_Opt_C...</td>\n",
              "      <td>3274668</td>\n",
              "      <td>296048</td>\n",
              "      <td>0.090406</td>\n",
              "      <td>0.356713</td>\n",
              "      <td>3754</td>\n",
              "      <td>0.005172</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4</th>\n",
              "      <td>dataset-downloader-kit/CV/dataset/Town01_Opt_C...</td>\n",
              "      <td>3278316</td>\n",
              "      <td>286544</td>\n",
              "      <td>0.087406</td>\n",
              "      <td>0.392556</td>\n",
              "      <td>565</td>\n",
              "      <td>0.005175</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>...</th>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>91</th>\n",
              "      <td>dataset-downloader-kit/CV/dataset/Town01_Opt_C...</td>\n",
              "      <td>3484380</td>\n",
              "      <td>337264</td>\n",
              "      <td>0.096793</td>\n",
              "      <td>0.458859</td>\n",
              "      <td>3561</td>\n",
              "      <td>0.005114</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>92</th>\n",
              "      <td>dataset-downloader-kit/CV/dataset/Town01_Opt_C...</td>\n",
              "      <td>3407268</td>\n",
              "      <td>319548</td>\n",
              "      <td>0.093784</td>\n",
              "      <td>0.544925</td>\n",
              "      <td>433</td>\n",
              "      <td>0.004996</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>93</th>\n",
              "      <td>dataset-downloader-kit/CV/dataset/Town01_Opt_C...</td>\n",
              "      <td>3227700</td>\n",
              "      <td>309704</td>\n",
              "      <td>0.095952</td>\n",
              "      <td>0.411857</td>\n",
              "      <td>603</td>\n",
              "      <td>0.005946</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>94</th>\n",
              "      <td>dataset-downloader-kit/CV/dataset/Town01_Opt_C...</td>\n",
              "      <td>3256320</td>\n",
              "      <td>298220</td>\n",
              "      <td>0.091582</td>\n",
              "      <td>0.419531</td>\n",
              "      <td>306</td>\n",
              "      <td>0.005488</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>95</th>\n",
              "      <td>dataset-downloader-kit/CV/dataset/Town01_Opt_C...</td>\n",
              "      <td>3288924</td>\n",
              "      <td>283980</td>\n",
              "      <td>0.086344</td>\n",
              "      <td>0.502103</td>\n",
              "      <td>467</td>\n",
              "      <td>0.005999</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "<p>96 rows × 7 columns</p>\n",
              "</div>"
            ],
            "text/plain": [
              "                                             filename  original_size_bytes  \\\n",
              "0   dataset-downloader-kit/CV/dataset/Town01_Opt_C...              3416088   \n",
              "1   dataset-downloader-kit/CV/dataset/Town01_Opt_C...              3251676   \n",
              "2   dataset-downloader-kit/CV/dataset/Town01_Opt_C...              3462324   \n",
              "3   dataset-downloader-kit/CV/dataset/Town01_Opt_C...              3274668   \n",
              "4   dataset-downloader-kit/CV/dataset/Town01_Opt_C...              3278316   \n",
              "..                                                ...                  ...   \n",
              "91  dataset-downloader-kit/CV/dataset/Town01_Opt_C...              3484380   \n",
              "92  dataset-downloader-kit/CV/dataset/Town01_Opt_C...              3407268   \n",
              "93  dataset-downloader-kit/CV/dataset/Town01_Opt_C...              3227700   \n",
              "94  dataset-downloader-kit/CV/dataset/Town01_Opt_C...              3256320   \n",
              "95  dataset-downloader-kit/CV/dataset/Town01_Opt_C...              3288924   \n",
              "\n",
              "    encoded_size  compression_ratio  encoding_decoding_time  lost_points  \\\n",
              "0         314512           0.092068                0.615055            0   \n",
              "1         299772           0.092190                0.452345            0   \n",
              "2         334056           0.096483                0.349651         1316   \n",
              "3         296048           0.090406                0.356713         3754   \n",
              "4         286544           0.087406                0.392556          565   \n",
              "..           ...                ...                     ...          ...   \n",
              "91        337264           0.096793                0.458859         3561   \n",
              "92        319548           0.093784                0.544925          433   \n",
              "93        309704           0.095952                0.411857          603   \n",
              "94        298220           0.091582                0.419531          306   \n",
              "95        283980           0.086344                0.502103          467   \n",
              "\n",
              "        mpde  \n",
              "0   0.005322  \n",
              "1   0.005561  \n",
              "2   0.006069  \n",
              "3   0.005172  \n",
              "4   0.005175  \n",
              "..       ...  \n",
              "91  0.005114  \n",
              "92  0.004996  \n",
              "93  0.005946  \n",
              "94  0.005488  \n",
              "95  0.005999  \n",
              "\n",
              "[96 rows x 7 columns]"
            ]
          },
          "execution_count": 3,
          "metadata": {},
          "output_type": "execute_result"
        }
      ],
      "source": [
        "draco_unlabeled = pd.read_csv('results/draco_unlabeled.csv')\n",
        "\n",
        "try:    \n",
        "    draco_unlabeled = draco_unlabeled.drop(columns=['encoded_size_bytes', 'compression_ratio_bytes', 'original_size'])\n",
        "except:\n",
        "    pass\n",
        "\n",
        "draco_unlabeled"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 9,
      "metadata": {},
      "outputs": [
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            " Average compression ratio: 8.9 % pm 0.5 %\n",
            " Average encoding-decoding time: 0.42 seconds pm 0.07 seconds\n",
            " Average lost points: 1861.33 points pm 1578.30 points\n",
            " Average MPDE: 0.005557 meters pm 0.0005 meters\n"
          ]
        }
      ],
      "source": [
        "# plot some results\n",
        "import matplotlib.pyplot as plt\n",
        "import seaborn as sns\n",
        "sns.set_theme(style='whitegrid', font_scale=1.5)\n",
        "\n",
        "\n",
        "# aggregate the results\n",
        "stats = pd.DataFrame(columns=draco_unlabeled.columns)\n",
        "stats.loc['avg'] = draco_unlabeled.mean()\n",
        "stats.loc['std'] = draco_unlabeled.std()\n",
        "\n",
        "print (f\" Average compression ratio: {stats.loc['avg', 'compression_ratio'] * 100:.1f} % pm {stats.loc['std', 'compression_ratio'] * 100:.1f} %\")\n",
        "print (f\" Average encoding-decoding time: {stats.loc['avg', 'encoding_decoding_time']:.2f} seconds pm {stats.loc['std', 'encoding_decoding_time']:.2f} seconds\")\n",
        "print (f\" Average lost points: {stats.loc['avg', 'lost_points']:.2f} points pm {stats.loc['std', 'lost_points']:.2f} points\")\n",
        "print (f\" Average MPDE: {stats.loc['avg', 'mpde']:4f} meters pm {stats.loc['std', 'mpde']:.4f} meters\")"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 10,
      "metadata": {},
      "outputs": [
        {
          "data": {
            "text/plain": [
              "<matplotlib.legend.Legend at 0x139392490>"
            ]
          },
          "execution_count": 10,
          "metadata": {},
          "output_type": "execute_result"
        },
        {
          "data": {
            "image/png": "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",
            "text/plain": [
              "<Figure size 1440x432 with 2 Axes>"
            ]
          },
          "metadata": {
            "needs_background": "light"
          },
          "output_type": "display_data"
        }
      ],
      "source": [
        "# plot \n",
        "sns.set_theme(style='whitegrid', font_scale=1.5, palette='Dark2')\n",
        "fig, ax = plt.subplots(1, 2, figsize=(20, 6))\n",
        "sns.histplot(data=draco_unlabeled, x='compression_ratio', ax=ax[0], bins=20)\n",
        "sns.histplot(data=draco_unlabeled, x='encoding_decoding_time', ax=ax[1], bins=20) \n",
        "\n",
        "# add mean and std\n",
        "lw = 4\n",
        "ax[0].set_title('Compression ratio')\n",
        "ax[1].set_title('Encoding-decoding time')\n",
        "\n",
        "ax[0].set_xlabel('Compression ratio')\n",
        "ax[1].set_xlabel('Encoding-decoding time')\n",
        "\n",
        "\n",
        "ax[0].axvline(stats.loc['avg', 'compression_ratio'], color='black', linestyle='--', label='mean', linewidth=lw)\n",
        "ax[0].axvline(stats.loc['avg', 'compression_ratio'] + stats.loc['std', 'compression_ratio'], color='black', linestyle=':', label='std', linewidth=lw)\n",
        "ax[0].axvline(stats.loc['avg', 'compression_ratio'] - stats.loc['std', 'compression_ratio'], color='black', linestyle=':', linewidth=lw)\n",
        "ax[0].legend()\n",
        "\n",
        "ax[1].axvline(stats.loc['avg', 'encoding_decoding_time'], color='black', linestyle='--', label='mean', linewidth=lw)\n",
        "ax[1].axvline(stats.loc['avg', 'encoding_decoding_time'] + stats.loc['std', 'encoding_decoding_time'], color='black', linestyle=':', label='std', linewidth=lw)\n",
        "ax[1].axvline(stats.loc['avg', 'encoding_decoding_time'] - stats.loc['std', 'encoding_decoding_time'], color='black', linestyle=':', linewidth=lw)\n",
        "ax[1].legend()\n",
        "\n"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 11,
      "metadata": {},
      "outputs": [
        {
          "data": {
            "image/png": "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",
            "text/plain": [
              "<Figure size 864x432 with 2 Axes>"
            ]
          },
          "metadata": {},
          "output_type": "display_data"
        }
      ],
      "source": [
        "# quality vs compression \n",
        "sns.set_theme(style='whitegrid', font_scale=1.5, palette='deep')\n",
        "\n",
        "fig, ax = plt.subplots(1, 2, figsize=(12, 6))\n",
        "sns.scatterplot(data=draco_unlabeled, x='compression_ratio', y='mpde', ax=ax[0])\n",
        "sns.scatterplot(data=draco_unlabeled, x='encoding_decoding_time', y='mpde', ax=ax[1])\n",
        "ax[0].set_title('Compression ratio vs MPDE')\n",
        "ax[1].set_title('Encoding-decoding time vs MPDE')\n",
        "ax[0].set_xlabel('Compression ratio')\n",
        "ax[1].set_xlabel('Encoding-decoding time')\n",
        "ax[0].set_ylabel('MPDE')\n",
        "ax[1].set_ylabel('')\n",
        "ax[1].set_yticklabels([])\n",
        "# ax[0].set_xticks(np.arange(0.\n",
        "# share axis \n",
        "ax[0].set_ylim(ax[1].get_ylim())\n",
        "fig.subplots_adjust(wspace=0.05)\n",
        "\n"
      ]
    },
    {
      "attachments": {},
      "cell_type": "markdown",
      "metadata": {
        "id": "_W6MuAWudfu8"
      },
      "source": [
        "## Labeled pointcloud\n",
        "\n",
        "Let us now consider the dataset with the corresponding labels. In this case we compress the point cloud obtained by compressing each label individually and we compute all the considered metrics."
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 11,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "-Ni_jkSIdf7m",
        "outputId": "010912c8-fc15-46ef-b5da-deca81b0acc0"
      },
      "outputs": [],
      "source": [
        "def draco_labeled (pc_file):\n",
        "    # Read the ply file\n",
        "    plydata = plyfile.PlyData.read(pc_file)\n",
        "\n",
        "    # Extract the XYZ coordinates of the points\n",
        "    points = np.column_stack((plydata['vertex']['x'], plydata['vertex']['y'], plydata['vertex']['z']))\n",
        "    # Extract the labels of the points\n",
        "    labels = plydata['vertex']['ObjTag']\n",
        "    print (f\"Number of labels: {len(np.unique(labels))}\")\n",
        "\n",
        "    # iterate over the labels and encode each label separately\n",
        "\n",
        "    results_PC = pd.DataFrame(columns=['label',             \n",
        "                                        'original_size',\n",
        "                                        'encoded_size', \n",
        "                                        'compression_ratio',  \n",
        "                                        'encoding_decoding_time', \n",
        "                                        'lost_points', 'mpde'])\n",
        "\n",
        "    for label in np.unique(labels):\n",
        "        start = time.time()\n",
        "        mask = labels == label\n",
        "\n",
        "        pc_label = points[mask]\n",
        "        # print (f\"Number of points in label {label}: {len(pc_label)}\") \n",
        "\n",
        "        # Encode the point cloud\n",
        "        encoded_data = DracoPy.encode(pc_label)\n",
        "\n",
        "        # Measure the size of the encoded data\n",
        "        encoded_size = len(encoded_data)\n",
        "        encoded_size_bytes = len(encoded_data)\n",
        "        original_size_bytes = sys.getsizeof(pc_label.tobytes())\n",
        "\n",
        "        # Measure the number of points in the original point cloud\n",
        "        original_size = len(points)\n",
        "        print (f\"Original size: {original_size_bytes} bytes \\t Encoded size: {encoded_size_bytes} bytes\")\n",
        "        # Decode the encoded data\n",
        "        decoded_points = DracoPy.decode(encoded_data)\n",
        "\n",
        "        mpde = mean_point_distance_error(pc_label, decoded_points.points)\n",
        "\n",
        "        # End the timer\n",
        "        end = time.time()\n",
        "\n",
        "        # Calculate the compression ratio\n",
        "        compression_ratio_bytes =  (encoded_size_bytes / original_size_bytes)\n",
        "\n",
        "        # Calculate the number of lost points\n",
        "        lost_points = len(points) - len(decoded_points.points)\n",
        "        results_PC = results_PC.append({'label': label,     \n",
        "                                        'original_size': original_size,         \n",
        "                                        'encoded_size': encoded_size,\n",
        "                                        'compression_ratio': compression_ratio_bytes,\n",
        "                                        'encoding_decoding_time': end - start,\n",
        "                                        'lost_points': lost_points,\n",
        "                                        'mpde': mpde\n",
        "                                        },\n",
        "                                        ignore_index=True)\n",
        "\n",
        "    # save the results\n",
        "    img_name = pc_file.split('/')[-1].split('.')[0]\n",
        "    res_dir = 'results'\n",
        "    \n",
        "    results_PC.to_csv(os.path.join(res_dir, img_name+'draco_labeled.csv'), index=False)\n",
        "\n",
        "    return results_PC\n",
        "        "
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 12,
      "metadata": {},
      "outputs": [
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Number of labels: 18\n",
            "Original size: 537 bytes \t Encoded size: 256 bytes\n",
            "Original size: 535365 bytes \t Encoded size: 91892 bytes\n",
            "Original size: 7221 bytes \t Encoded size: 1780 bytes\n",
            "Original size: 11049 bytes \t Encoded size: 1988 bytes\n",
            "Original size: 1082925 bytes \t Encoded size: 102120 bytes\n",
            "Original size: 699273 bytes \t Encoded size: 95308 bytes\n",
            "Original size: 4353 bytes \t Encoded size: 1268 bytes\n",
            "Original size: 562653 bytes \t Encoded size: 70192 bytes\n",
            "Original size: 138417 bytes \t Encoded size: 25412 bytes\n",
            "Original size: 81 bytes \t Encoded size: 96 bytes\n",
            "Original size: 1149 bytes \t Encoded size: 356 bytes\n",
            "Original size: 321 bytes \t Encoded size: 152 bytes\n",
            "Original size: 83997 bytes \t Encoded size: 12940 bytes\n",
            "Original size: 46125 bytes \t Encoded size: 7016 bytes\n",
            "Original size: 31857 bytes \t Encoded size: 5144 bytes\n",
            "Original size: 196389 bytes \t Encoded size: 16640 bytes\n",
            "Original size: 13821 bytes \t Encoded size: 3768 bytes\n",
            "Original size: 1149 bytes \t Encoded size: 400 bytes\n",
            "Number of labels: 20\n",
            "Original size: 64101 bytes \t Encoded size: 14424 bytes\n",
            "Original size: 202629 bytes \t Encoded size: 39520 bytes\n",
            "Original size: 21741 bytes \t Encoded size: 4716 bytes\n",
            "Original size: 369 bytes \t Encoded size: 200 bytes\n",
            "Original size: 16149 bytes \t Encoded size: 3148 bytes\n",
            "Original size: 1186353 bytes \t Encoded size: 102576 bytes\n",
            "Original size: 721269 bytes \t Encoded size: 89860 bytes\n",
            "Original size: 175485 bytes \t Encoded size: 37444 bytes\n",
            "Original size: 562653 bytes \t Encoded size: 70176 bytes\n",
            "Original size: 11709 bytes \t Encoded size: 2416 bytes\n",
            "Original size: 33333 bytes \t Encoded size: 6696 bytes\n",
            "Original size: 2889 bytes \t Encoded size: 660 bytes\n",
            "Original size: 67581 bytes \t Encoded size: 8744 bytes\n",
            "Original size: 729 bytes \t Encoded size: 244 bytes\n",
            "Original size: 117717 bytes \t Encoded size: 17020 bytes\n",
            "Original size: 7185 bytes \t Encoded size: 1320 bytes\n",
            "Original size: 52269 bytes \t Encoded size: 8464 bytes\n",
            "Original size: 585 bytes \t Encoded size: 288 bytes\n",
            "Original size: 3393 bytes \t Encoded size: 1080 bytes\n",
            "Original size: 4197 bytes \t Encoded size: 884 bytes\n",
            "Number of labels: 18\n",
            "Original size: 211401 bytes \t Encoded size: 40036 bytes\n",
            "Original size: 340629 bytes \t Encoded size: 66312 bytes\n",
            "Original size: 52185 bytes \t Encoded size: 8984 bytes\n",
            "Original size: 11589 bytes \t Encoded size: 2384 bytes\n",
            "Original size: 1195629 bytes \t Encoded size: 109940 bytes\n",
            "Original size: 669009 bytes \t Encoded size: 83864 bytes\n",
            "Original size: 293661 bytes \t Encoded size: 57504 bytes\n",
            "Original size: 562653 bytes \t Encoded size: 70184 bytes\n",
            "Original size: 51861 bytes \t Encoded size: 9976 bytes\n",
            "Original size: 489 bytes \t Encoded size: 220 bytes\n",
            "Original size: 957 bytes \t Encoded size: 336 bytes\n",
            "Original size: 22053 bytes \t Encoded size: 3520 bytes\n",
            "Original size: 15093 bytes \t Encoded size: 3148 bytes\n",
            "Original size: 4773 bytes \t Encoded size: 956 bytes\n",
            "Original size: 24837 bytes \t Encoded size: 4768 bytes\n",
            "Original size: 2841 bytes \t Encoded size: 1052 bytes\n",
            "Original size: 45 bytes \t Encoded size: 80 bytes\n",
            "Original size: 3213 bytes \t Encoded size: 736 bytes\n",
            "Number of labels: 21\n",
            "Original size: 77877 bytes \t Encoded size: 15480 bytes\n",
            "Original size: 251421 bytes \t Encoded size: 50648 bytes\n",
            "Original size: 9561 bytes \t Encoded size: 2060 bytes\n",
            "Original size: 93 bytes \t Encoded size: 104 bytes\n",
            "Original size: 18681 bytes \t Encoded size: 3148 bytes\n",
            "Original size: 1180617 bytes \t Encoded size: 105796 bytes\n",
            "Original size: 590289 bytes \t Encoded size: 72304 bytes\n",
            "Original size: 101229 bytes \t Encoded size: 19588 bytes\n",
            "Original size: 562653 bytes \t Encoded size: 70192 bytes\n",
            "Original size: 99261 bytes \t Encoded size: 19592 bytes\n",
            "Original size: 105 bytes \t Encoded size: 104 bytes\n",
            "Original size: 45 bytes \t Encoded size: 80 bytes\n",
            "Original size: 165 bytes \t Encoded size: 120 bytes\n",
            "Original size: 2469 bytes \t Encoded size: 544 bytes\n",
            "Original size: 56229 bytes \t Encoded size: 8472 bytes\n",
            "Original size: 15717 bytes \t Encoded size: 3332 bytes\n",
            "Original size: 61605 bytes \t Encoded size: 9616 bytes\n",
            "Original size: 16389 bytes \t Encoded size: 2136 bytes\n",
            "Original size: 214509 bytes \t Encoded size: 27308 bytes\n",
            "Original size: 645 bytes \t Encoded size: 312 bytes\n",
            "Original size: 15801 bytes \t Encoded size: 2296 bytes\n",
            "Number of labels: 20\n",
            "Original size: 392217 bytes \t Encoded size: 61000 bytes\n",
            "Original size: 31113 bytes \t Encoded size: 6500 bytes\n",
            "Original size: 11685 bytes \t Encoded size: 2448 bytes\n",
            "Original size: 1317081 bytes \t Encoded size: 123092 bytes\n",
            "Original size: 443181 bytes \t Encoded size: 54644 bytes\n",
            "Original size: 100509 bytes \t Encoded size: 21564 bytes\n",
            "Original size: 562653 bytes \t Encoded size: 70196 bytes\n",
            "Original size: 7245 bytes \t Encoded size: 1960 bytes\n",
            "Original size: 1593 bytes \t Encoded size: 472 bytes\n",
            "Original size: 41241 bytes \t Encoded size: 6776 bytes\n",
            "Original size: 1545 bytes \t Encoded size: 456 bytes\n",
            "Original size: 5157 bytes \t Encoded size: 1104 bytes\n",
            "Original size: 669 bytes \t Encoded size: 264 bytes\n",
            "Original size: 33213 bytes \t Encoded size: 4656 bytes\n",
            "Original size: 25017 bytes \t Encoded size: 4168 bytes\n",
            "Original size: 267189 bytes \t Encoded size: 32812 bytes\n",
            "Original size: 20469 bytes \t Encoded size: 4632 bytes\n",
            "Original size: 849 bytes \t Encoded size: 356 bytes\n",
            "Original size: 153 bytes \t Encoded size: 120 bytes\n",
            "Original size: 16197 bytes \t Encoded size: 2844 bytes\n",
            "Number of labels: 18\n",
            "Original size: 535689 bytes \t Encoded size: 77500 bytes\n",
            "Original size: 30777 bytes \t Encoded size: 7864 bytes\n",
            "Original size: 77013 bytes \t Encoded size: 14116 bytes\n",
            "Original size: 14577 bytes \t Encoded size: 2836 bytes\n",
            "Original size: 1150269 bytes \t Encoded size: 106404 bytes\n",
            "Original size: 597165 bytes \t Encoded size: 71324 bytes\n",
            "Original size: 91353 bytes \t Encoded size: 21100 bytes\n",
            "Original size: 562653 bytes \t Encoded size: 70180 bytes\n",
            "Original size: 21429 bytes \t Encoded size: 4844 bytes\n",
            "Original size: 1029 bytes \t Encoded size: 412 bytes\n",
            "Original size: 2769 bytes \t Encoded size: 616 bytes\n",
            "Original size: 7401 bytes \t Encoded size: 2008 bytes\n",
            "Original size: 13041 bytes \t Encoded size: 3120 bytes\n",
            "Original size: 75657 bytes \t Encoded size: 10704 bytes\n",
            "Original size: 2685 bytes \t Encoded size: 652 bytes\n",
            "Original size: 129705 bytes \t Encoded size: 19024 bytes\n",
            "Original size: 19869 bytes \t Encoded size: 4276 bytes\n",
            "Original size: 1317 bytes \t Encoded size: 392 bytes\n",
            "Number of labels: 18\n",
            "Original size: 134301 bytes \t Encoded size: 24600 bytes\n",
            "Original size: 251313 bytes \t Encoded size: 46528 bytes\n",
            "Original size: 25749 bytes \t Encoded size: 5408 bytes\n",
            "Original size: 14169 bytes \t Encoded size: 2700 bytes\n",
            "Original size: 1202301 bytes \t Encoded size: 116024 bytes\n",
            "Original size: 757989 bytes \t Encoded size: 100836 bytes\n",
            "Original size: 301197 bytes \t Encoded size: 62320 bytes\n",
            "Original size: 562653 bytes \t Encoded size: 70188 bytes\n",
            "Original size: 88257 bytes \t Encoded size: 16588 bytes\n",
            "Original size: 3465 bytes \t Encoded size: 756 bytes\n",
            "Original size: 2157 bytes \t Encoded size: 760 bytes\n",
            "Original size: 42633 bytes \t Encoded size: 8960 bytes\n",
            "Original size: 1389 bytes \t Encoded size: 416 bytes\n",
            "Original size: 49281 bytes \t Encoded size: 8128 bytes\n",
            "Original size: 3069 bytes \t Encoded size: 656 bytes\n",
            "Original size: 13929 bytes \t Encoded size: 2748 bytes\n",
            "Original size: 45 bytes \t Encoded size: 80 bytes\n",
            "Original size: 1089 bytes \t Encoded size: 336 bytes\n",
            "Number of labels: 18\n",
            "Original size: 392337 bytes \t Encoded size: 72652 bytes\n",
            "Original size: 15681 bytes \t Encoded size: 4712 bytes\n",
            "Original size: 216477 bytes \t Encoded size: 35800 bytes\n",
            "Original size: 12225 bytes \t Encoded size: 2448 bytes\n",
            "Original size: 1310445 bytes \t Encoded size: 139776 bytes\n",
            "Original size: 501633 bytes \t Encoded size: 72820 bytes\n",
            "Original size: 86841 bytes \t Encoded size: 18744 bytes\n",
            "Original size: 562641 bytes \t Encoded size: 70208 bytes\n",
            "Original size: 225 bytes \t Encoded size: 148 bytes\n",
            "Original size: 3825 bytes \t Encoded size: 932 bytes\n",
            "Original size: 45 bytes \t Encoded size: 80 bytes\n",
            "Original size: 39633 bytes \t Encoded size: 6004 bytes\n",
            "Original size: 141 bytes \t Encoded size: 112 bytes\n",
            "Original size: 405 bytes \t Encoded size: 204 bytes\n",
            "Original size: 43197 bytes \t Encoded size: 7584 bytes\n",
            "Original size: 129 bytes \t Encoded size: 112 bytes\n",
            "Original size: 113817 bytes \t Encoded size: 15920 bytes\n",
            "Original size: 57 bytes \t Encoded size: 88 bytes\n",
            "Number of labels: 20\n",
            "Original size: 105549 bytes \t Encoded size: 18988 bytes\n",
            "Original size: 214281 bytes \t Encoded size: 35608 bytes\n",
            "Original size: 23493 bytes \t Encoded size: 4452 bytes\n",
            "Original size: 8145 bytes \t Encoded size: 1808 bytes\n",
            "Original size: 1166829 bytes \t Encoded size: 103720 bytes\n",
            "Original size: 700845 bytes \t Encoded size: 86596 bytes\n",
            "Original size: 271077 bytes \t Encoded size: 57944 bytes\n",
            "Original size: 562653 bytes \t Encoded size: 70188 bytes\n",
            "Original size: 23457 bytes \t Encoded size: 4684 bytes\n",
            "Original size: 45 bytes \t Encoded size: 80 bytes\n",
            "Original size: 15945 bytes \t Encoded size: 3728 bytes\n",
            "Original size: 57 bytes \t Encoded size: 88 bytes\n",
            "Original size: 1461 bytes \t Encoded size: 472 bytes\n",
            "Original size: 909 bytes \t Encoded size: 316 bytes\n",
            "Original size: 2277 bytes \t Encoded size: 824 bytes\n",
            "Original size: 88245 bytes \t Encoded size: 10836 bytes\n",
            "Original size: 3585 bytes \t Encoded size: 716 bytes\n",
            "Original size: 98349 bytes \t Encoded size: 12404 bytes\n",
            "Original size: 69 bytes \t Encoded size: 92 bytes\n",
            "Original size: 2049 bytes \t Encoded size: 512 bytes\n",
            "Number of labels: 20\n",
            "Original size: 61377 bytes \t Encoded size: 14484 bytes\n",
            "Original size: 352005 bytes \t Encoded size: 65032 bytes\n",
            "Original size: 25425 bytes \t Encoded size: 5280 bytes\n",
            "Original size: 69 bytes \t Encoded size: 92 bytes\n",
            "Original size: 7821 bytes \t Encoded size: 1776 bytes\n",
            "Original size: 1124457 bytes \t Encoded size: 104452 bytes\n",
            "Original size: 711645 bytes \t Encoded size: 89168 bytes\n",
            "Original size: 44733 bytes \t Encoded size: 8848 bytes\n",
            "Original size: 562653 bytes \t Encoded size: 70200 bytes\n",
            "Original size: 2553 bytes \t Encoded size: 792 bytes\n",
            "Original size: 19329 bytes \t Encoded size: 4252 bytes\n",
            "Original size: 297 bytes \t Encoded size: 168 bytes\n",
            "Original size: 70233 bytes \t Encoded size: 8216 bytes\n",
            "Original size: 765 bytes \t Encoded size: 264 bytes\n",
            "Original size: 48717 bytes \t Encoded size: 7624 bytes\n",
            "Original size: 9429 bytes \t Encoded size: 1588 bytes\n",
            "Original size: 185373 bytes \t Encoded size: 24080 bytes\n",
            "Original size: 10473 bytes \t Encoded size: 2672 bytes\n",
            "Original size: 105 bytes \t Encoded size: 104 bytes\n",
            "Original size: 5709 bytes \t Encoded size: 1856 bytes\n",
            "Number of labels: 21\n",
            "Original size: 41169 bytes \t Encoded size: 7940 bytes\n",
            "Original size: 233061 bytes \t Encoded size: 40644 bytes\n",
            "Original size: 67953 bytes \t Encoded size: 10240 bytes\n",
            "Original size: 5517 bytes \t Encoded size: 1204 bytes\n",
            "Original size: 16881 bytes \t Encoded size: 2548 bytes\n",
            "Original size: 1158537 bytes \t Encoded size: 103560 bytes\n",
            "Original size: 609477 bytes \t Encoded size: 70640 bytes\n",
            "Original size: 44109 bytes \t Encoded size: 9864 bytes\n",
            "Original size: 562653 bytes \t Encoded size: 70208 bytes\n",
            "Original size: 24177 bytes \t Encoded size: 4704 bytes\n",
            "Original size: 4917 bytes \t Encoded size: 1488 bytes\n",
            "Original size: 5721 bytes \t Encoded size: 1384 bytes\n",
            "Original size: 1569 bytes \t Encoded size: 436 bytes\n",
            "Original size: 5421 bytes \t Encoded size: 1260 bytes\n",
            "Original size: 93 bytes \t Encoded size: 96 bytes\n",
            "Original size: 51645 bytes \t Encoded size: 6828 bytes\n",
            "Original size: 1101 bytes \t Encoded size: 320 bytes\n",
            "Original size: 113301 bytes \t Encoded size: 14964 bytes\n",
            "Original size: 21009 bytes \t Encoded size: 6492 bytes\n",
            "Original size: 2157 bytes \t Encoded size: 780 bytes\n",
            "Original size: 477 bytes \t Encoded size: 196 bytes\n",
            "Number of labels: 19\n",
            "Original size: 107421 bytes \t Encoded size: 21072 bytes\n",
            "Original size: 127941 bytes \t Encoded size: 27336 bytes\n",
            "Original size: 88101 bytes \t Encoded size: 10948 bytes\n",
            "Original size: 26805 bytes \t Encoded size: 3852 bytes\n",
            "Original size: 1236405 bytes \t Encoded size: 110968 bytes\n",
            "Original size: 592365 bytes \t Encoded size: 74544 bytes\n",
            "Original size: 195249 bytes \t Encoded size: 38400 bytes\n",
            "Original size: 562653 bytes \t Encoded size: 70180 bytes\n",
            "Original size: 55401 bytes \t Encoded size: 10780 bytes\n",
            "Original size: 105 bytes \t Encoded size: 104 bytes\n",
            "Original size: 14289 bytes \t Encoded size: 2636 bytes\n",
            "Original size: 2229 bytes \t Encoded size: 612 bytes\n",
            "Original size: 24765 bytes \t Encoded size: 4160 bytes\n",
            "Original size: 201 bytes \t Encoded size: 136 bytes\n",
            "Original size: 46509 bytes \t Encoded size: 7964 bytes\n",
            "Original size: 6417 bytes \t Encoded size: 1508 bytes\n",
            "Original size: 161769 bytes \t Encoded size: 23216 bytes\n",
            "Original size: 249 bytes \t Encoded size: 144 bytes\n",
            "Original size: 8817 bytes \t Encoded size: 2308 bytes\n",
            "Number of labels: 19\n",
            "Original size: 83001 bytes \t Encoded size: 17284 bytes\n",
            "Original size: 171357 bytes \t Encoded size: 32188 bytes\n",
            "Original size: 44133 bytes \t Encoded size: 7920 bytes\n",
            "Original size: 59181 bytes \t Encoded size: 4324 bytes\n",
            "Original size: 1203489 bytes \t Encoded size: 102824 bytes\n",
            "Original size: 625785 bytes \t Encoded size: 75256 bytes\n",
            "Original size: 162105 bytes \t Encoded size: 37092 bytes\n",
            "Original size: 562653 bytes \t Encoded size: 70188 bytes\n",
            "Original size: 197961 bytes \t Encoded size: 30792 bytes\n",
            "Original size: 28593 bytes \t Encoded size: 6360 bytes\n",
            "Original size: 165 bytes \t Encoded size: 128 bytes\n",
            "Original size: 33537 bytes \t Encoded size: 5472 bytes\n",
            "Original size: 4833 bytes \t Encoded size: 892 bytes\n",
            "Original size: 1869 bytes \t Encoded size: 596 bytes\n",
            "Original size: 22785 bytes \t Encoded size: 4468 bytes\n",
            "Original size: 321 bytes \t Encoded size: 188 bytes\n",
            "Original size: 10497 bytes \t Encoded size: 2124 bytes\n",
            "Original size: 9741 bytes \t Encoded size: 2800 bytes\n",
            "Original size: 141 bytes \t Encoded size: 116 bytes\n",
            "Number of labels: 18\n",
            "Original size: 143325 bytes \t Encoded size: 27840 bytes\n",
            "Original size: 24045 bytes \t Encoded size: 6012 bytes\n",
            "Original size: 72573 bytes \t Encoded size: 13288 bytes\n",
            "Original size: 14685 bytes \t Encoded size: 2368 bytes\n",
            "Original size: 1395417 bytes \t Encoded size: 133904 bytes\n",
            "Original size: 270717 bytes \t Encoded size: 36600 bytes\n",
            "Original size: 113589 bytes \t Encoded size: 26684 bytes\n",
            "Original size: 562653 bytes \t Encoded size: 70220 bytes\n",
            "Original size: 8685 bytes \t Encoded size: 2160 bytes\n",
            "Original size: 429 bytes \t Encoded size: 188 bytes\n",
            "Original size: 148701 bytes \t Encoded size: 21260 bytes\n",
            "Original size: 1773 bytes \t Encoded size: 512 bytes\n",
            "Original size: 1917 bytes \t Encoded size: 600 bytes\n",
            "Original size: 4905 bytes \t Encoded size: 1448 bytes\n",
            "Original size: 29721 bytes \t Encoded size: 5408 bytes\n",
            "Original size: 4989 bytes \t Encoded size: 992 bytes\n",
            "Original size: 300981 bytes \t Encoded size: 39080 bytes\n",
            "Original size: 1665 bytes \t Encoded size: 480 bytes\n",
            "Number of labels: 19\n",
            "Original size: 68301 bytes \t Encoded size: 14520 bytes\n",
            "Original size: 310017 bytes \t Encoded size: 57572 bytes\n",
            "Original size: 35625 bytes \t Encoded size: 7584 bytes\n",
            "Original size: 381 bytes \t Encoded size: 212 bytes\n",
            "Original size: 24801 bytes \t Encoded size: 3504 bytes\n",
            "Original size: 1305033 bytes \t Encoded size: 133860 bytes\n",
            "Original size: 558249 bytes \t Encoded size: 70308 bytes\n",
            "Original size: 263853 bytes \t Encoded size: 46372 bytes\n",
            "Original size: 562653 bytes \t Encoded size: 70180 bytes\n",
            "Original size: 68385 bytes \t Encoded size: 11464 bytes\n",
            "Original size: 1689 bytes \t Encoded size: 592 bytes\n",
            "Original size: 2541 bytes \t Encoded size: 604 bytes\n",
            "Original size: 32337 bytes \t Encoded size: 4336 bytes\n",
            "Original size: 20865 bytes \t Encoded size: 4920 bytes\n",
            "Original size: 77601 bytes \t Encoded size: 10996 bytes\n",
            "Original size: 1497 bytes \t Encoded size: 460 bytes\n",
            "Original size: 15885 bytes \t Encoded size: 2868 bytes\n",
            "Original size: 3429 bytes \t Encoded size: 1132 bytes\n",
            "Original size: 897 bytes \t Encoded size: 324 bytes\n",
            "Number of labels: 16\n",
            "Original size: 353889 bytes \t Encoded size: 60864 bytes\n",
            "Original size: 176373 bytes \t Encoded size: 34328 bytes\n",
            "Original size: 11565 bytes \t Encoded size: 2568 bytes\n",
            "Original size: 10881 bytes \t Encoded size: 2088 bytes\n",
            "Original size: 1160325 bytes \t Encoded size: 107536 bytes\n",
            "Original size: 642153 bytes \t Encoded size: 78996 bytes\n",
            "Original size: 227277 bytes \t Encoded size: 49688 bytes\n",
            "Original size: 562653 bytes \t Encoded size: 70188 bytes\n",
            "Original size: 69177 bytes \t Encoded size: 13024 bytes\n",
            "Original size: 477 bytes \t Encoded size: 240 bytes\n",
            "Original size: 249 bytes \t Encoded size: 136 bytes\n",
            "Original size: 81 bytes \t Encoded size: 96 bytes\n",
            "Original size: 17877 bytes \t Encoded size: 3324 bytes\n",
            "Original size: 753 bytes \t Encoded size: 276 bytes\n",
            "Original size: 78165 bytes \t Encoded size: 11340 bytes\n",
            "Original size: 381 bytes \t Encoded size: 176 bytes\n",
            "Number of labels: 19\n",
            "Original size: 4017 bytes \t Encoded size: 1136 bytes\n",
            "Original size: 623745 bytes \t Encoded size: 102504 bytes\n",
            "Original size: 4401 bytes \t Encoded size: 1252 bytes\n",
            "Original size: 29793 bytes \t Encoded size: 3844 bytes\n",
            "Original size: 1198089 bytes \t Encoded size: 100808 bytes\n",
            "Original size: 748953 bytes \t Encoded size: 95796 bytes\n",
            "Original size: 12837 bytes \t Encoded size: 3860 bytes\n",
            "Original size: 562653 bytes \t Encoded size: 70192 bytes\n",
            "Original size: 39909 bytes \t Encoded size: 7280 bytes\n",
            "Original size: 1857 bytes \t Encoded size: 620 bytes\n",
            "Original size: 1761 bytes \t Encoded size: 528 bytes\n",
            "Original size: 9513 bytes \t Encoded size: 2016 bytes\n",
            "Original size: 7569 bytes \t Encoded size: 1472 bytes\n",
            "Original size: 67917 bytes \t Encoded size: 9344 bytes\n",
            "Original size: 825 bytes \t Encoded size: 284 bytes\n",
            "Original size: 30117 bytes \t Encoded size: 4884 bytes\n",
            "Original size: 4569 bytes \t Encoded size: 1540 bytes\n",
            "Original size: 8541 bytes \t Encoded size: 1708 bytes\n",
            "Original size: 741 bytes \t Encoded size: 288 bytes\n",
            "Number of labels: 18\n",
            "Original size: 28317 bytes \t Encoded size: 5768 bytes\n",
            "Original size: 179601 bytes \t Encoded size: 32064 bytes\n",
            "Original size: 37209 bytes \t Encoded size: 6952 bytes\n",
            "Original size: 11157 bytes \t Encoded size: 2008 bytes\n",
            "Original size: 1608045 bytes \t Encoded size: 140260 bytes\n",
            "Original size: 226341 bytes \t Encoded size: 35876 bytes\n",
            "Original size: 62409 bytes \t Encoded size: 10716 bytes\n",
            "Original size: 562653 bytes \t Encoded size: 70184 bytes\n",
            "Original size: 36441 bytes \t Encoded size: 7168 bytes\n",
            "Original size: 73293 bytes \t Encoded size: 15704 bytes\n",
            "Original size: 2781 bytes \t Encoded size: 696 bytes\n",
            "Original size: 11721 bytes \t Encoded size: 2192 bytes\n",
            "Original size: 1221 bytes \t Encoded size: 428 bytes\n",
            "Original size: 39933 bytes \t Encoded size: 6248 bytes\n",
            "Original size: 321 bytes \t Encoded size: 160 bytes\n",
            "Original size: 169257 bytes \t Encoded size: 22956 bytes\n",
            "Original size: 525 bytes \t Encoded size: 220 bytes\n",
            "Original size: 477 bytes \t Encoded size: 196 bytes\n",
            "Number of labels: 19\n",
            "Original size: 57225 bytes \t Encoded size: 13656 bytes\n",
            "Original size: 350685 bytes \t Encoded size: 58692 bytes\n",
            "Original size: 27525 bytes \t Encoded size: 5320 bytes\n",
            "Original size: 62493 bytes \t Encoded size: 3816 bytes\n",
            "Original size: 1148421 bytes \t Encoded size: 107488 bytes\n",
            "Original size: 710157 bytes \t Encoded size: 93140 bytes\n",
            "Original size: 37101 bytes \t Encoded size: 7544 bytes\n",
            "Original size: 562653 bytes \t Encoded size: 70188 bytes\n",
            "Original size: 4221 bytes \t Encoded size: 1204 bytes\n",
            "Original size: 20229 bytes \t Encoded size: 4368 bytes\n",
            "Original size: 45 bytes \t Encoded size: 80 bytes\n",
            "Original size: 17829 bytes \t Encoded size: 3392 bytes\n",
            "Original size: 1341 bytes \t Encoded size: 412 bytes\n",
            "Original size: 54057 bytes \t Encoded size: 8016 bytes\n",
            "Original size: 153 bytes \t Encoded size: 120 bytes\n",
            "Original size: 41433 bytes \t Encoded size: 6732 bytes\n",
            "Original size: 100965 bytes \t Encoded size: 14176 bytes\n",
            "Original size: 24957 bytes \t Encoded size: 5692 bytes\n",
            "Original size: 141 bytes \t Encoded size: 112 bytes\n",
            "Number of labels: 19\n",
            "Original size: 117573 bytes \t Encoded size: 23188 bytes\n",
            "Original size: 247149 bytes \t Encoded size: 49052 bytes\n",
            "Original size: 69897 bytes \t Encoded size: 13136 bytes\n",
            "Original size: 25005 bytes \t Encoded size: 3288 bytes\n",
            "Original size: 1141965 bytes \t Encoded size: 99520 bytes\n",
            "Original size: 646785 bytes \t Encoded size: 79672 bytes\n",
            "Original size: 138105 bytes \t Encoded size: 28476 bytes\n",
            "Original size: 562653 bytes \t Encoded size: 70180 bytes\n",
            "Original size: 79593 bytes \t Encoded size: 12392 bytes\n",
            "Original size: 45 bytes \t Encoded size: 80 bytes\n",
            "Original size: 645 bytes \t Encoded size: 248 bytes\n",
            "Original size: 7665 bytes \t Encoded size: 1720 bytes\n",
            "Original size: 85281 bytes \t Encoded size: 12600 bytes\n",
            "Original size: 34989 bytes \t Encoded size: 5232 bytes\n",
            "Original size: 3489 bytes \t Encoded size: 920 bytes\n",
            "Original size: 110337 bytes \t Encoded size: 13892 bytes\n",
            "Original size: 48777 bytes \t Encoded size: 8180 bytes\n",
            "Original size: 45969 bytes \t Encoded size: 8656 bytes\n",
            "Original size: 1113 bytes \t Encoded size: 388 bytes\n",
            "Number of labels: 18\n",
            "Original size: 29889 bytes \t Encoded size: 7224 bytes\n",
            "Original size: 413397 bytes \t Encoded size: 67344 bytes\n",
            "Original size: 14361 bytes \t Encoded size: 3224 bytes\n",
            "Original size: 13881 bytes \t Encoded size: 2740 bytes\n",
            "Original size: 1188933 bytes \t Encoded size: 100696 bytes\n",
            "Original size: 789249 bytes \t Encoded size: 100000 bytes\n",
            "Original size: 37113 bytes \t Encoded size: 7948 bytes\n",
            "Original size: 562653 bytes \t Encoded size: 70196 bytes\n",
            "Original size: 3417 bytes \t Encoded size: 972 bytes\n",
            "Original size: 4341 bytes \t Encoded size: 1164 bytes\n",
            "Original size: 23109 bytes \t Encoded size: 4068 bytes\n",
            "Original size: 4677 bytes \t Encoded size: 1056 bytes\n",
            "Original size: 56181 bytes \t Encoded size: 7744 bytes\n",
            "Original size: 165 bytes \t Encoded size: 124 bytes\n",
            "Original size: 38301 bytes \t Encoded size: 6956 bytes\n",
            "Original size: 2205 bytes \t Encoded size: 828 bytes\n",
            "Original size: 20157 bytes \t Encoded size: 4460 bytes\n",
            "Original size: 93 bytes \t Encoded size: 100 bytes\n",
            "Number of labels: 19\n",
            "Original size: 325377 bytes \t Encoded size: 53208 bytes\n",
            "Original size: 59277 bytes \t Encoded size: 13280 bytes\n",
            "Original size: 22665 bytes \t Encoded size: 4180 bytes\n",
            "Original size: 15741 bytes \t Encoded size: 2820 bytes\n",
            "Original size: 1217865 bytes \t Encoded size: 104972 bytes\n",
            "Original size: 627357 bytes \t Encoded size: 73652 bytes\n",
            "Original size: 179025 bytes \t Encoded size: 39144 bytes\n",
            "Original size: 562653 bytes \t Encoded size: 70184 bytes\n",
            "Original size: 6489 bytes \t Encoded size: 1732 bytes\n",
            "Original size: 297 bytes \t Encoded size: 180 bytes\n",
            "Original size: 161853 bytes \t Encoded size: 28596 bytes\n",
            "Original size: 2505 bytes \t Encoded size: 540 bytes\n",
            "Original size: 2745 bytes \t Encoded size: 824 bytes\n",
            "Original size: 489 bytes \t Encoded size: 196 bytes\n",
            "Original size: 717 bytes \t Encoded size: 324 bytes\n",
            "Original size: 41097 bytes \t Encoded size: 6332 bytes\n",
            "Original size: 669 bytes \t Encoded size: 264 bytes\n",
            "Original size: 54033 bytes \t Encoded size: 8516 bytes\n",
            "Original size: 393 bytes \t Encoded size: 192 bytes\n",
            "Number of labels: 19\n",
            "Original size: 1641 bytes \t Encoded size: 500 bytes\n",
            "Original size: 152109 bytes \t Encoded size: 28304 bytes\n",
            "Original size: 67137 bytes \t Encoded size: 9920 bytes\n",
            "Original size: 12009 bytes \t Encoded size: 1964 bytes\n",
            "Original size: 1123605 bytes \t Encoded size: 108088 bytes\n",
            "Original size: 630945 bytes \t Encoded size: 76836 bytes\n",
            "Original size: 110517 bytes \t Encoded size: 16988 bytes\n",
            "Original size: 562653 bytes \t Encoded size: 70184 bytes\n",
            "Original size: 33993 bytes \t Encoded size: 6088 bytes\n",
            "Original size: 13101 bytes \t Encoded size: 3384 bytes\n",
            "Original size: 2001 bytes \t Encoded size: 464 bytes\n",
            "Original size: 10605 bytes \t Encoded size: 1940 bytes\n",
            "Original size: 2649 bytes \t Encoded size: 916 bytes\n",
            "Original size: 47745 bytes \t Encoded size: 6104 bytes\n",
            "Original size: 18393 bytes \t Encoded size: 3292 bytes\n",
            "Original size: 94029 bytes \t Encoded size: 15080 bytes\n",
            "Original size: 92337 bytes \t Encoded size: 23260 bytes\n",
            "Original size: 225 bytes \t Encoded size: 148 bytes\n",
            "Original size: 5553 bytes \t Encoded size: 1156 bytes\n",
            "Number of labels: 22\n",
            "Original size: 121185 bytes \t Encoded size: 20700 bytes\n",
            "Original size: 190989 bytes \t Encoded size: 35288 bytes\n",
            "Original size: 7605 bytes \t Encoded size: 1956 bytes\n",
            "Original size: 213 bytes \t Encoded size: 140 bytes\n",
            "Original size: 24153 bytes \t Encoded size: 3672 bytes\n",
            "Original size: 1180881 bytes \t Encoded size: 99384 bytes\n",
            "Original size: 666333 bytes \t Encoded size: 79376 bytes\n",
            "Original size: 118365 bytes \t Encoded size: 24136 bytes\n",
            "Original size: 562653 bytes \t Encoded size: 70188 bytes\n",
            "Original size: 130209 bytes \t Encoded size: 22388 bytes\n",
            "Original size: 1977 bytes \t Encoded size: 700 bytes\n",
            "Original size: 2613 bytes \t Encoded size: 836 bytes\n",
            "Original size: 2469 bytes \t Encoded size: 548 bytes\n",
            "Original size: 18537 bytes \t Encoded size: 3920 bytes\n",
            "Original size: 13713 bytes \t Encoded size: 2568 bytes\n",
            "Original size: 84741 bytes \t Encoded size: 10532 bytes\n",
            "Original size: 897 bytes \t Encoded size: 308 bytes\n",
            "Original size: 91905 bytes \t Encoded size: 13496 bytes\n",
            "Original size: 2037 bytes \t Encoded size: 740 bytes\n",
            "Original size: 1089 bytes \t Encoded size: 424 bytes\n",
            "Original size: 57 bytes \t Encoded size: 88 bytes\n",
            "Original size: 6177 bytes \t Encoded size: 1560 bytes\n",
            "Number of labels: 21\n",
            "Original size: 185241 bytes \t Encoded size: 31448 bytes\n",
            "Original size: 49665 bytes \t Encoded size: 11312 bytes\n",
            "Original size: 4713 bytes \t Encoded size: 1292 bytes\n",
            "Original size: 45 bytes \t Encoded size: 80 bytes\n",
            "Original size: 12237 bytes \t Encoded size: 2536 bytes\n",
            "Original size: 1160001 bytes \t Encoded size: 101188 bytes\n",
            "Original size: 633321 bytes \t Encoded size: 74180 bytes\n",
            "Original size: 205005 bytes \t Encoded size: 42316 bytes\n",
            "Original size: 562653 bytes \t Encoded size: 70180 bytes\n",
            "Original size: 3165 bytes \t Encoded size: 936 bytes\n",
            "Original size: 177 bytes \t Encoded size: 132 bytes\n",
            "Original size: 310461 bytes \t Encoded size: 53248 bytes\n",
            "Original size: 261 bytes \t Encoded size: 164 bytes\n",
            "Original size: 2025 bytes \t Encoded size: 496 bytes\n",
            "Original size: 1365 bytes \t Encoded size: 468 bytes\n",
            "Original size: 201 bytes \t Encoded size: 144 bytes\n",
            "Original size: 1761 bytes \t Encoded size: 660 bytes\n",
            "Original size: 9897 bytes \t Encoded size: 2296 bytes\n",
            "Original size: 5937 bytes \t Encoded size: 1136 bytes\n",
            "Original size: 105033 bytes \t Encoded size: 14948 bytes\n",
            "Original size: 3573 bytes \t Encoded size: 768 bytes\n",
            "Number of labels: 19\n",
            "Original size: 357 bytes \t Encoded size: 172 bytes\n",
            "Original size: 362421 bytes \t Encoded size: 59068 bytes\n",
            "Original size: 41265 bytes \t Encoded size: 7860 bytes\n",
            "Original size: 14589 bytes \t Encoded size: 2848 bytes\n",
            "Original size: 1593429 bytes \t Encoded size: 142324 bytes\n",
            "Original size: 244773 bytes \t Encoded size: 37320 bytes\n",
            "Original size: 115689 bytes \t Encoded size: 23580 bytes\n",
            "Original size: 562653 bytes \t Encoded size: 70224 bytes\n",
            "Original size: 15969 bytes \t Encoded size: 3972 bytes\n",
            "Original size: 1305 bytes \t Encoded size: 388 bytes\n",
            "Original size: 61017 bytes \t Encoded size: 8944 bytes\n",
            "Original size: 2121 bytes \t Encoded size: 616 bytes\n",
            "Original size: 13845 bytes \t Encoded size: 2828 bytes\n",
            "Original size: 525 bytes \t Encoded size: 232 bytes\n",
            "Original size: 31725 bytes \t Encoded size: 5396 bytes\n",
            "Original size: 6165 bytes \t Encoded size: 2024 bytes\n",
            "Original size: 146361 bytes \t Encoded size: 20052 bytes\n",
            "Original size: 2049 bytes \t Encoded size: 744 bytes\n",
            "Original size: 5157 bytes \t Encoded size: 1000 bytes\n",
            "Number of labels: 19\n",
            "Original size: 162201 bytes \t Encoded size: 30976 bytes\n",
            "Original size: 29589 bytes \t Encoded size: 7168 bytes\n",
            "Original size: 84633 bytes \t Encoded size: 14608 bytes\n",
            "Original size: 12621 bytes \t Encoded size: 2608 bytes\n",
            "Original size: 1441317 bytes \t Encoded size: 127156 bytes\n",
            "Original size: 277449 bytes \t Encoded size: 39060 bytes\n",
            "Original size: 134097 bytes \t Encoded size: 31200 bytes\n",
            "Original size: 562653 bytes \t Encoded size: 70220 bytes\n",
            "Original size: 14877 bytes \t Encoded size: 3164 bytes\n",
            "Original size: 789 bytes \t Encoded size: 268 bytes\n",
            "Original size: 36897 bytes \t Encoded size: 6472 bytes\n",
            "Original size: 3309 bytes \t Encoded size: 804 bytes\n",
            "Original size: 4857 bytes \t Encoded size: 1332 bytes\n",
            "Original size: 5733 bytes \t Encoded size: 1676 bytes\n",
            "Original size: 66705 bytes \t Encoded size: 9204 bytes\n",
            "Original size: 5253 bytes \t Encoded size: 1104 bytes\n",
            "Original size: 364293 bytes \t Encoded size: 43172 bytes\n",
            "Original size: 6273 bytes \t Encoded size: 1592 bytes\n",
            "Original size: 201 bytes \t Encoded size: 132 bytes\n",
            "Number of labels: 18\n",
            "Original size: 382437 bytes \t Encoded size: 65732 bytes\n",
            "Original size: 54033 bytes \t Encoded size: 11564 bytes\n",
            "Original size: 837 bytes \t Encoded size: 296 bytes\n",
            "Original size: 21789 bytes \t Encoded size: 3144 bytes\n",
            "Original size: 1194189 bytes \t Encoded size: 102468 bytes\n",
            "Original size: 642309 bytes \t Encoded size: 71020 bytes\n",
            "Original size: 241461 bytes \t Encoded size: 50120 bytes\n",
            "Original size: 562653 bytes \t Encoded size: 70180 bytes\n",
            "Original size: 58389 bytes \t Encoded size: 12176 bytes\n",
            "Original size: 357 bytes \t Encoded size: 156 bytes\n",
            "Original size: 621 bytes \t Encoded size: 292 bytes\n",
            "Original size: 45 bytes \t Encoded size: 80 bytes\n",
            "Original size: 3189 bytes \t Encoded size: 752 bytes\n",
            "Original size: 273 bytes \t Encoded size: 176 bytes\n",
            "Original size: 55593 bytes \t Encoded size: 7424 bytes\n",
            "Original size: 969 bytes \t Encoded size: 320 bytes\n",
            "Original size: 62877 bytes \t Encoded size: 9516 bytes\n",
            "Original size: 309 bytes \t Encoded size: 164 bytes\n",
            "Number of labels: 16\n",
            "Original size: 194385 bytes \t Encoded size: 38092 bytes\n",
            "Original size: 17985 bytes \t Encoded size: 5032 bytes\n",
            "Original size: 97533 bytes \t Encoded size: 24596 bytes\n",
            "Original size: 1809 bytes \t Encoded size: 536 bytes\n",
            "Original size: 908121 bytes \t Encoded size: 95056 bytes\n",
            "Original size: 520065 bytes \t Encoded size: 69188 bytes\n",
            "Original size: 200889 bytes \t Encoded size: 42932 bytes\n",
            "Original size: 562653 bytes \t Encoded size: 70188 bytes\n",
            "Original size: 7137 bytes \t Encoded size: 1704 bytes\n",
            "Original size: 1449 bytes \t Encoded size: 540 bytes\n",
            "Original size: 501 bytes \t Encoded size: 236 bytes\n",
            "Original size: 73581 bytes \t Encoded size: 10040 bytes\n",
            "Original size: 6753 bytes \t Encoded size: 1872 bytes\n",
            "Original size: 247917 bytes \t Encoded size: 25316 bytes\n",
            "Original size: 614769 bytes \t Encoded size: 122140 bytes\n",
            "Original size: 3549 bytes \t Encoded size: 1076 bytes\n",
            "Number of labels: 18\n",
            "Original size: 304665 bytes \t Encoded size: 47528 bytes\n",
            "Original size: 40317 bytes \t Encoded size: 9136 bytes\n",
            "Original size: 5553 bytes \t Encoded size: 1448 bytes\n",
            "Original size: 15405 bytes \t Encoded size: 2852 bytes\n",
            "Original size: 1187349 bytes \t Encoded size: 101260 bytes\n",
            "Original size: 615933 bytes \t Encoded size: 74344 bytes\n",
            "Original size: 205845 bytes \t Encoded size: 42572 bytes\n",
            "Original size: 562653 bytes \t Encoded size: 70180 bytes\n",
            "Original size: 40641 bytes \t Encoded size: 7076 bytes\n",
            "Original size: 2193 bytes \t Encoded size: 760 bytes\n",
            "Original size: 20265 bytes \t Encoded size: 4840 bytes\n",
            "Original size: 45 bytes \t Encoded size: 80 bytes\n",
            "Original size: 31413 bytes \t Encoded size: 5532 bytes\n",
            "Original size: 123789 bytes \t Encoded size: 12676 bytes\n",
            "Original size: 15321 bytes \t Encoded size: 1996 bytes\n",
            "Original size: 55341 bytes \t Encoded size: 8464 bytes\n",
            "Original size: 6897 bytes \t Encoded size: 2072 bytes\n",
            "Original size: 15489 bytes \t Encoded size: 2256 bytes\n",
            "Number of labels: 20\n",
            "Original size: 577533 bytes \t Encoded size: 81368 bytes\n",
            "Original size: 19701 bytes \t Encoded size: 4548 bytes\n",
            "Original size: 2505 bytes \t Encoded size: 736 bytes\n",
            "Original size: 18885 bytes \t Encoded size: 3164 bytes\n",
            "Original size: 1157961 bytes \t Encoded size: 103336 bytes\n",
            "Original size: 601641 bytes \t Encoded size: 69040 bytes\n",
            "Original size: 132765 bytes \t Encoded size: 28968 bytes\n",
            "Original size: 562653 bytes \t Encoded size: 70172 bytes\n",
            "Original size: 13449 bytes \t Encoded size: 3048 bytes\n",
            "Original size: 2745 bytes \t Encoded size: 892 bytes\n",
            "Original size: 4125 bytes \t Encoded size: 1328 bytes\n",
            "Original size: 2133 bytes \t Encoded size: 464 bytes\n",
            "Original size: 861 bytes \t Encoded size: 340 bytes\n",
            "Original size: 1713 bytes \t Encoded size: 396 bytes\n",
            "Original size: 39885 bytes \t Encoded size: 5908 bytes\n",
            "Original size: 8529 bytes \t Encoded size: 1824 bytes\n",
            "Original size: 188577 bytes \t Encoded size: 24284 bytes\n",
            "Original size: 453 bytes \t Encoded size: 236 bytes\n",
            "Original size: 2349 bytes \t Encoded size: 600 bytes\n",
            "Original size: 4857 bytes \t Encoded size: 1084 bytes\n",
            "Number of labels: 18\n",
            "Original size: 5109 bytes \t Encoded size: 1392 bytes\n",
            "Original size: 492189 bytes \t Encoded size: 68524 bytes\n",
            "Original size: 48285 bytes \t Encoded size: 8736 bytes\n",
            "Original size: 10629 bytes \t Encoded size: 1896 bytes\n",
            "Original size: 1169085 bytes \t Encoded size: 99392 bytes\n",
            "Original size: 590061 bytes \t Encoded size: 69132 bytes\n",
            "Original size: 80901 bytes \t Encoded size: 17500 bytes\n",
            "Original size: 562641 bytes \t Encoded size: 70188 bytes\n",
            "Original size: 669 bytes \t Encoded size: 272 bytes\n",
            "Original size: 57 bytes \t Encoded size: 88 bytes\n",
            "Original size: 57 bytes \t Encoded size: 88 bytes\n",
            "Original size: 100989 bytes \t Encoded size: 9148 bytes\n",
            "Original size: 1593 bytes \t Encoded size: 452 bytes\n",
            "Original size: 40353 bytes \t Encoded size: 5936 bytes\n",
            "Original size: 2793 bytes \t Encoded size: 748 bytes\n",
            "Original size: 91341 bytes \t Encoded size: 10616 bytes\n",
            "Original size: 2361 bytes \t Encoded size: 864 bytes\n",
            "Original size: 741 bytes \t Encoded size: 276 bytes\n",
            "Number of labels: 22\n",
            "Original size: 78873 bytes \t Encoded size: 14136 bytes\n",
            "Original size: 171069 bytes \t Encoded size: 33636 bytes\n",
            "Original size: 10233 bytes \t Encoded size: 2556 bytes\n",
            "Original size: 717 bytes \t Encoded size: 300 bytes\n",
            "Original size: 16629 bytes \t Encoded size: 2992 bytes\n",
            "Original size: 1167141 bytes \t Encoded size: 103240 bytes\n",
            "Original size: 633249 bytes \t Encoded size: 74632 bytes\n",
            "Original size: 89445 bytes \t Encoded size: 15776 bytes\n",
            "Original size: 562653 bytes \t Encoded size: 70188 bytes\n",
            "Original size: 44205 bytes \t Encoded size: 8468 bytes\n",
            "Original size: 849 bytes \t Encoded size: 380 bytes\n",
            "Original size: 8709 bytes \t Encoded size: 2308 bytes\n",
            "Original size: 2421 bytes \t Encoded size: 504 bytes\n",
            "Original size: 89061 bytes \t Encoded size: 11380 bytes\n",
            "Original size: 3357 bytes \t Encoded size: 804 bytes\n",
            "Original size: 37713 bytes \t Encoded size: 5908 bytes\n",
            "Original size: 6249 bytes \t Encoded size: 1072 bytes\n",
            "Original size: 238281 bytes \t Encoded size: 28900 bytes\n",
            "Original size: 2925 bytes \t Encoded size: 772 bytes\n",
            "Original size: 3525 bytes \t Encoded size: 1176 bytes\n",
            "Original size: 225 bytes \t Encoded size: 148 bytes\n",
            "Original size: 5133 bytes \t Encoded size: 1124 bytes\n",
            "Number of labels: 18\n",
            "Original size: 503841 bytes \t Encoded size: 71696 bytes\n",
            "Original size: 22629 bytes \t Encoded size: 5540 bytes\n",
            "Original size: 4221 bytes \t Encoded size: 1128 bytes\n",
            "Original size: 8553 bytes \t Encoded size: 1960 bytes\n",
            "Original size: 1052805 bytes \t Encoded size: 88580 bytes\n",
            "Original size: 537393 bytes \t Encoded size: 60408 bytes\n",
            "Original size: 116073 bytes \t Encoded size: 26480 bytes\n",
            "Original size: 562653 bytes \t Encoded size: 70188 bytes\n",
            "Original size: 6057 bytes \t Encoded size: 1520 bytes\n",
            "Original size: 825 bytes \t Encoded size: 352 bytes\n",
            "Original size: 12021 bytes \t Encoded size: 3272 bytes\n",
            "Original size: 1401 bytes \t Encoded size: 388 bytes\n",
            "Original size: 1197 bytes \t Encoded size: 424 bytes\n",
            "Original size: 393 bytes \t Encoded size: 164 bytes\n",
            "Original size: 36753 bytes \t Encoded size: 6104 bytes\n",
            "Original size: 4701 bytes \t Encoded size: 984 bytes\n",
            "Original size: 500013 bytes \t Encoded size: 42336 bytes\n",
            "Original size: 2073 bytes \t Encoded size: 536 bytes\n",
            "Number of labels: 19\n",
            "Original size: 207321 bytes \t Encoded size: 36684 bytes\n",
            "Original size: 155133 bytes \t Encoded size: 34256 bytes\n",
            "Original size: 19233 bytes \t Encoded size: 4172 bytes\n",
            "Original size: 9465 bytes \t Encoded size: 1968 bytes\n",
            "Original size: 1169709 bytes \t Encoded size: 106768 bytes\n",
            "Original size: 542949 bytes \t Encoded size: 69508 bytes\n",
            "Original size: 197649 bytes \t Encoded size: 36156 bytes\n",
            "Original size: 562653 bytes \t Encoded size: 70180 bytes\n",
            "Original size: 59661 bytes \t Encoded size: 12056 bytes\n",
            "Original size: 105 bytes \t Encoded size: 100 bytes\n",
            "Original size: 2181 bytes \t Encoded size: 552 bytes\n",
            "Original size: 2865 bytes \t Encoded size: 756 bytes\n",
            "Original size: 5649 bytes \t Encoded size: 1424 bytes\n",
            "Original size: 345 bytes \t Encoded size: 172 bytes\n",
            "Original size: 32421 bytes \t Encoded size: 6172 bytes\n",
            "Original size: 4809 bytes \t Encoded size: 1000 bytes\n",
            "Original size: 314721 bytes \t Encoded size: 34204 bytes\n",
            "Original size: 921 bytes \t Encoded size: 392 bytes\n",
            "Original size: 6165 bytes \t Encoded size: 1584 bytes\n",
            "Number of labels: 19\n",
            "Original size: 45561 bytes \t Encoded size: 10476 bytes\n",
            "Original size: 362085 bytes \t Encoded size: 60544 bytes\n",
            "Original size: 26073 bytes \t Encoded size: 4976 bytes\n",
            "Original size: 189 bytes \t Encoded size: 136 bytes\n",
            "Original size: 15681 bytes \t Encoded size: 2968 bytes\n",
            "Original size: 1160673 bytes \t Encoded size: 96572 bytes\n",
            "Original size: 710901 bytes \t Encoded size: 87916 bytes\n",
            "Original size: 38025 bytes \t Encoded size: 8276 bytes\n",
            "Original size: 562653 bytes \t Encoded size: 70188 bytes\n",
            "Original size: 3141 bytes \t Encoded size: 1024 bytes\n",
            "Original size: 19713 bytes \t Encoded size: 4824 bytes\n",
            "Original size: 72525 bytes \t Encoded size: 7564 bytes\n",
            "Original size: 873 bytes \t Encoded size: 288 bytes\n",
            "Original size: 45489 bytes \t Encoded size: 7348 bytes\n",
            "Original size: 1977 bytes \t Encoded size: 544 bytes\n",
            "Original size: 97629 bytes \t Encoded size: 12688 bytes\n",
            "Original size: 873 bytes \t Encoded size: 384 bytes\n",
            "Original size: 46833 bytes \t Encoded size: 9748 bytes\n",
            "Original size: 741 bytes \t Encoded size: 296 bytes\n",
            "Number of labels: 16\n",
            "Original size: 411753 bytes \t Encoded size: 76452 bytes\n",
            "Original size: 633 bytes \t Encoded size: 292 bytes\n",
            "Original size: 382293 bytes \t Encoded size: 64660 bytes\n",
            "Original size: 9201 bytes \t Encoded size: 1788 bytes\n",
            "Original size: 1377285 bytes \t Encoded size: 136560 bytes\n",
            "Original size: 422637 bytes \t Encoded size: 62216 bytes\n",
            "Original size: 44841 bytes \t Encoded size: 10552 bytes\n",
            "Original size: 562641 bytes \t Encoded size: 70176 bytes\n",
            "Original size: 429 bytes \t Encoded size: 224 bytes\n",
            "Original size: 861 bytes \t Encoded size: 316 bytes\n",
            "Original size: 111429 bytes \t Encoded size: 20296 bytes\n",
            "Original size: 141 bytes \t Encoded size: 116 bytes\n",
            "Original size: 45 bytes \t Encoded size: 80 bytes\n",
            "Original size: 24081 bytes \t Encoded size: 3768 bytes\n",
            "Original size: 45 bytes \t Encoded size: 80 bytes\n",
            "Original size: 99585 bytes \t Encoded size: 16272 bytes\n",
            "Number of labels: 20\n",
            "Original size: 38025 bytes \t Encoded size: 8612 bytes\n",
            "Original size: 253665 bytes \t Encoded size: 49984 bytes\n",
            "Original size: 7401 bytes \t Encoded size: 1844 bytes\n",
            "Original size: 405 bytes \t Encoded size: 216 bytes\n",
            "Original size: 18021 bytes \t Encoded size: 3084 bytes\n",
            "Original size: 1246077 bytes \t Encoded size: 118220 bytes\n",
            "Original size: 620037 bytes \t Encoded size: 76236 bytes\n",
            "Original size: 242241 bytes \t Encoded size: 33352 bytes\n",
            "Original size: 562653 bytes \t Encoded size: 70192 bytes\n",
            "Original size: 167133 bytes \t Encoded size: 28992 bytes\n",
            "Original size: 57 bytes \t Encoded size: 88 bytes\n",
            "Original size: 45 bytes \t Encoded size: 80 bytes\n",
            "Original size: 2205 bytes \t Encoded size: 696 bytes\n",
            "Original size: 3021 bytes \t Encoded size: 612 bytes\n",
            "Original size: 46665 bytes \t Encoded size: 7304 bytes\n",
            "Original size: 13905 bytes \t Encoded size: 3304 bytes\n",
            "Original size: 39621 bytes \t Encoded size: 7080 bytes\n",
            "Original size: 9369 bytes \t Encoded size: 2016 bytes\n",
            "Original size: 1293 bytes \t Encoded size: 520 bytes\n",
            "Original size: 333 bytes \t Encoded size: 192 bytes\n",
            "Number of labels: 19\n",
            "Original size: 6153 bytes \t Encoded size: 1456 bytes\n",
            "Original size: 212049 bytes \t Encoded size: 42976 bytes\n",
            "Original size: 51117 bytes \t Encoded size: 8328 bytes\n",
            "Original size: 23553 bytes \t Encoded size: 3368 bytes\n",
            "Original size: 1143129 bytes \t Encoded size: 95920 bytes\n",
            "Original size: 616953 bytes \t Encoded size: 70516 bytes\n",
            "Original size: 228969 bytes \t Encoded size: 45420 bytes\n",
            "Original size: 562641 bytes \t Encoded size: 70180 bytes\n",
            "Original size: 145053 bytes \t Encoded size: 24600 bytes\n",
            "Original size: 4725 bytes \t Encoded size: 1128 bytes\n",
            "Original size: 237 bytes \t Encoded size: 148 bytes\n",
            "Original size: 381 bytes \t Encoded size: 196 bytes\n",
            "Original size: 169713 bytes \t Encoded size: 21772 bytes\n",
            "Original size: 2949 bytes \t Encoded size: 644 bytes\n",
            "Original size: 72177 bytes \t Encoded size: 9756 bytes\n",
            "Original size: 933 bytes \t Encoded size: 300 bytes\n",
            "Original size: 70497 bytes \t Encoded size: 8976 bytes\n",
            "Original size: 81 bytes \t Encoded size: 100 bytes\n",
            "Original size: 11193 bytes \t Encoded size: 1980 bytes\n",
            "Number of labels: 20\n",
            "Original size: 42477 bytes \t Encoded size: 9708 bytes\n",
            "Original size: 145869 bytes \t Encoded size: 30424 bytes\n",
            "Original size: 7845 bytes \t Encoded size: 2040 bytes\n",
            "Original size: 285 bytes \t Encoded size: 168 bytes\n",
            "Original size: 17505 bytes \t Encoded size: 2748 bytes\n",
            "Original size: 1222809 bytes \t Encoded size: 104504 bytes\n",
            "Original size: 623625 bytes \t Encoded size: 79288 bytes\n",
            "Original size: 142737 bytes \t Encoded size: 30988 bytes\n",
            "Original size: 562653 bytes \t Encoded size: 70212 bytes\n",
            "Original size: 9969 bytes \t Encoded size: 1828 bytes\n",
            "Original size: 14949 bytes \t Encoded size: 3320 bytes\n",
            "Original size: 1881 bytes \t Encoded size: 396 bytes\n",
            "Original size: 52185 bytes \t Encoded size: 7048 bytes\n",
            "Original size: 1797 bytes \t Encoded size: 712 bytes\n",
            "Original size: 74889 bytes \t Encoded size: 11128 bytes\n",
            "Original size: 32829 bytes \t Encoded size: 4008 bytes\n",
            "Original size: 141933 bytes \t Encoded size: 19588 bytes\n",
            "Original size: 144249 bytes \t Encoded size: 37772 bytes\n",
            "Original size: 129 bytes \t Encoded size: 112 bytes\n",
            "Original size: 17709 bytes \t Encoded size: 2892 bytes\n",
            "Number of labels: 21\n",
            "Original size: 39777 bytes \t Encoded size: 8816 bytes\n",
            "Original size: 114597 bytes \t Encoded size: 23716 bytes\n",
            "Original size: 16737 bytes \t Encoded size: 3576 bytes\n",
            "Original size: 153 bytes \t Encoded size: 116 bytes\n",
            "Original size: 18177 bytes \t Encoded size: 3040 bytes\n",
            "Original size: 1431057 bytes \t Encoded size: 130596 bytes\n",
            "Original size: 445425 bytes \t Encoded size: 59004 bytes\n",
            "Original size: 99789 bytes \t Encoded size: 18676 bytes\n",
            "Original size: 562653 bytes \t Encoded size: 70184 bytes\n",
            "Original size: 35721 bytes \t Encoded size: 7176 bytes\n",
            "Original size: 1053 bytes \t Encoded size: 412 bytes\n",
            "Original size: 3225 bytes \t Encoded size: 932 bytes\n",
            "Original size: 2973 bytes \t Encoded size: 616 bytes\n",
            "Original size: 58761 bytes \t Encoded size: 9440 bytes\n",
            "Original size: 1461 bytes \t Encoded size: 476 bytes\n",
            "Original size: 35457 bytes \t Encoded size: 5668 bytes\n",
            "Original size: 3861 bytes \t Encoded size: 820 bytes\n",
            "Original size: 301053 bytes \t Encoded size: 33176 bytes\n",
            "Original size: 3681 bytes \t Encoded size: 1172 bytes\n",
            "Original size: 93 bytes \t Encoded size: 104 bytes\n",
            "Original size: 5685 bytes \t Encoded size: 1192 bytes\n",
            "Number of labels: 18\n",
            "Original size: 43077 bytes \t Encoded size: 9964 bytes\n",
            "Original size: 375921 bytes \t Encoded size: 63480 bytes\n",
            "Original size: 23817 bytes \t Encoded size: 4576 bytes\n",
            "Original size: 105 bytes \t Encoded size: 104 bytes\n",
            "Original size: 14457 bytes \t Encoded size: 2924 bytes\n",
            "Original size: 1179777 bytes \t Encoded size: 100212 bytes\n",
            "Original size: 764997 bytes \t Encoded size: 97308 bytes\n",
            "Original size: 40893 bytes \t Encoded size: 8972 bytes\n",
            "Original size: 562653 bytes \t Encoded size: 70208 bytes\n",
            "Original size: 5517 bytes \t Encoded size: 1500 bytes\n",
            "Original size: 17385 bytes \t Encoded size: 4268 bytes\n",
            "Original size: 52137 bytes \t Encoded size: 6964 bytes\n",
            "Original size: 2409 bytes \t Encoded size: 636 bytes\n",
            "Original size: 40581 bytes \t Encoded size: 6568 bytes\n",
            "Original size: 357 bytes \t Encoded size: 168 bytes\n",
            "Original size: 29493 bytes \t Encoded size: 4164 bytes\n",
            "Original size: 52401 bytes \t Encoded size: 10636 bytes\n",
            "Original size: 249 bytes \t Encoded size: 144 bytes\n",
            "Number of labels: 17\n",
            "Original size: 252897 bytes \t Encoded size: 38120 bytes\n",
            "Original size: 107517 bytes \t Encoded size: 20484 bytes\n",
            "Original size: 13305 bytes \t Encoded size: 2540 bytes\n",
            "Original size: 24477 bytes \t Encoded size: 2928 bytes\n",
            "Original size: 1102977 bytes \t Encoded size: 91256 bytes\n",
            "Original size: 562197 bytes \t Encoded size: 64840 bytes\n",
            "Original size: 272721 bytes \t Encoded size: 56468 bytes\n",
            "Original size: 562653 bytes \t Encoded size: 70176 bytes\n",
            "Original size: 13377 bytes \t Encoded size: 2404 bytes\n",
            "Original size: 45 bytes \t Encoded size: 80 bytes\n",
            "Original size: 2385 bytes \t Encoded size: 804 bytes\n",
            "Original size: 969 bytes \t Encoded size: 328 bytes\n",
            "Original size: 47997 bytes \t Encoded size: 6016 bytes\n",
            "Original size: 45 bytes \t Encoded size: 80 bytes\n",
            "Original size: 32745 bytes \t Encoded size: 5572 bytes\n",
            "Original size: 427809 bytes \t Encoded size: 48604 bytes\n",
            "Original size: 45 bytes \t Encoded size: 80 bytes\n",
            "Number of labels: 20\n",
            "Original size: 37689 bytes \t Encoded size: 8672 bytes\n",
            "Original size: 132381 bytes \t Encoded size: 28268 bytes\n",
            "Original size: 16905 bytes \t Encoded size: 3460 bytes\n",
            "Original size: 14409 bytes \t Encoded size: 2828 bytes\n",
            "Original size: 1164489 bytes \t Encoded size: 102524 bytes\n",
            "Original size: 638193 bytes \t Encoded size: 70240 bytes\n",
            "Original size: 101745 bytes \t Encoded size: 20728 bytes\n",
            "Original size: 562653 bytes \t Encoded size: 70208 bytes\n",
            "Original size: 291297 bytes \t Encoded size: 37024 bytes\n",
            "Original size: 14781 bytes \t Encoded size: 3580 bytes\n",
            "Original size: 45 bytes \t Encoded size: 80 bytes\n",
            "Original size: 5013 bytes \t Encoded size: 1284 bytes\n",
            "Original size: 2913 bytes \t Encoded size: 732 bytes\n",
            "Original size: 657 bytes \t Encoded size: 260 bytes\n",
            "Original size: 91317 bytes \t Encoded size: 10944 bytes\n",
            "Original size: 6765 bytes \t Encoded size: 1252 bytes\n",
            "Original size: 80373 bytes \t Encoded size: 11848 bytes\n",
            "Original size: 23685 bytes \t Encoded size: 5508 bytes\n",
            "Original size: 9957 bytes \t Encoded size: 2348 bytes\n",
            "Original size: 3477 bytes \t Encoded size: 748 bytes\n",
            "Number of labels: 20\n",
            "Original size: 40569 bytes \t Encoded size: 9876 bytes\n",
            "Original size: 370317 bytes \t Encoded size: 61668 bytes\n",
            "Original size: 12969 bytes \t Encoded size: 3032 bytes\n",
            "Original size: 69 bytes \t Encoded size: 92 bytes\n",
            "Original size: 10665 bytes \t Encoded size: 2104 bytes\n",
            "Original size: 805809 bytes \t Encoded size: 89520 bytes\n",
            "Original size: 715005 bytes \t Encoded size: 90340 bytes\n",
            "Original size: 83637 bytes \t Encoded size: 15900 bytes\n",
            "Original size: 562653 bytes \t Encoded size: 70196 bytes\n",
            "Original size: 4389 bytes \t Encoded size: 1236 bytes\n",
            "Original size: 45 bytes \t Encoded size: 80 bytes\n",
            "Original size: 11841 bytes \t Encoded size: 2936 bytes\n",
            "Original size: 249 bytes \t Encoded size: 152 bytes\n",
            "Original size: 20277 bytes \t Encoded size: 3844 bytes\n",
            "Original size: 81 bytes \t Encoded size: 92 bytes\n",
            "Original size: 89001 bytes \t Encoded size: 13688 bytes\n",
            "Original size: 261 bytes \t Encoded size: 164 bytes\n",
            "Original size: 571353 bytes \t Encoded size: 17480 bytes\n",
            "Original size: 6261 bytes \t Encoded size: 1896 bytes\n",
            "Original size: 129 bytes \t Encoded size: 112 bytes\n",
            "Number of labels: 19\n",
            "Original size: 79845 bytes \t Encoded size: 15872 bytes\n",
            "Original size: 150153 bytes \t Encoded size: 28004 bytes\n",
            "Original size: 14949 bytes \t Encoded size: 3260 bytes\n",
            "Original size: 24513 bytes \t Encoded size: 3136 bytes\n",
            "Original size: 1190313 bytes \t Encoded size: 102140 bytes\n",
            "Original size: 660021 bytes \t Encoded size: 79296 bytes\n",
            "Original size: 291813 bytes \t Encoded size: 61444 bytes\n",
            "Original size: 562653 bytes \t Encoded size: 70180 bytes\n",
            "Original size: 9285 bytes \t Encoded size: 2316 bytes\n",
            "Original size: 93 bytes \t Encoded size: 100 bytes\n",
            "Original size: 221949 bytes \t Encoded size: 34064 bytes\n",
            "Original size: 69 bytes \t Encoded size: 92 bytes\n",
            "Original size: 1545 bytes \t Encoded size: 496 bytes\n",
            "Original size: 525 bytes \t Encoded size: 228 bytes\n",
            "Original size: 4269 bytes \t Encoded size: 1436 bytes\n",
            "Original size: 29133 bytes \t Encoded size: 4636 bytes\n",
            "Original size: 981 bytes \t Encoded size: 416 bytes\n",
            "Original size: 8481 bytes \t Encoded size: 1872 bytes\n",
            "Original size: 489 bytes \t Encoded size: 208 bytes\n",
            "Number of labels: 20\n",
            "Original size: 53385 bytes \t Encoded size: 10612 bytes\n",
            "Original size: 194145 bytes \t Encoded size: 37436 bytes\n",
            "Original size: 8781 bytes \t Encoded size: 2304 bytes\n",
            "Original size: 2229 bytes \t Encoded size: 656 bytes\n",
            "Original size: 20937 bytes \t Encoded size: 3256 bytes\n",
            "Original size: 1223109 bytes \t Encoded size: 105912 bytes\n",
            "Original size: 728685 bytes \t Encoded size: 86520 bytes\n",
            "Original size: 94965 bytes \t Encoded size: 18188 bytes\n",
            "Original size: 562653 bytes \t Encoded size: 70188 bytes\n",
            "Original size: 91041 bytes \t Encoded size: 17244 bytes\n",
            "Original size: 3117 bytes \t Encoded size: 1140 bytes\n",
            "Original size: 4557 bytes \t Encoded size: 1352 bytes\n",
            "Original size: 2673 bytes \t Encoded size: 572 bytes\n",
            "Original size: 36213 bytes \t Encoded size: 6688 bytes\n",
            "Original size: 5073 bytes \t Encoded size: 1156 bytes\n",
            "Original size: 84573 bytes \t Encoded size: 11312 bytes\n",
            "Original size: 177 bytes \t Encoded size: 120 bytes\n",
            "Original size: 66873 bytes \t Encoded size: 10552 bytes\n",
            "Original size: 2877 bytes \t Encoded size: 960 bytes\n",
            "Original size: 10209 bytes \t Encoded size: 1936 bytes\n",
            "Number of labels: 18\n",
            "Original size: 319617 bytes \t Encoded size: 48328 bytes\n",
            "Original size: 30729 bytes \t Encoded size: 7096 bytes\n",
            "Original size: 7101 bytes \t Encoded size: 1768 bytes\n",
            "Original size: 14661 bytes \t Encoded size: 2592 bytes\n",
            "Original size: 1067997 bytes \t Encoded size: 93584 bytes\n",
            "Original size: 593577 bytes \t Encoded size: 69876 bytes\n",
            "Original size: 188685 bytes \t Encoded size: 39228 bytes\n",
            "Original size: 562653 bytes \t Encoded size: 70188 bytes\n",
            "Original size: 23673 bytes \t Encoded size: 4520 bytes\n",
            "Original size: 2253 bytes \t Encoded size: 776 bytes\n",
            "Original size: 20757 bytes \t Encoded size: 4800 bytes\n",
            "Original size: 27825 bytes \t Encoded size: 4924 bytes\n",
            "Original size: 83673 bytes \t Encoded size: 8580 bytes\n",
            "Original size: 201 bytes \t Encoded size: 132 bytes\n",
            "Original size: 336597 bytes \t Encoded size: 30348 bytes\n",
            "Original size: 4257 bytes \t Encoded size: 1028 bytes\n",
            "Original size: 237 bytes \t Encoded size: 156 bytes\n",
            "Original size: 3885 bytes \t Encoded size: 1140 bytes\n",
            "Number of labels: 19\n",
            "Original size: 62301 bytes \t Encoded size: 13140 bytes\n",
            "Original size: 127845 bytes \t Encoded size: 28132 bytes\n",
            "Original size: 58185 bytes \t Encoded size: 9988 bytes\n",
            "Original size: 19749 bytes \t Encoded size: 3496 bytes\n",
            "Original size: 1226517 bytes \t Encoded size: 108452 bytes\n",
            "Original size: 462813 bytes \t Encoded size: 54724 bytes\n",
            "Original size: 297189 bytes \t Encoded size: 58080 bytes\n",
            "Original size: 562653 bytes \t Encoded size: 70180 bytes\n",
            "Original size: 29001 bytes \t Encoded size: 6472 bytes\n",
            "Original size: 345 bytes \t Encoded size: 152 bytes\n",
            "Original size: 241629 bytes \t Encoded size: 25448 bytes\n",
            "Original size: 2181 bytes \t Encoded size: 604 bytes\n",
            "Original size: 8685 bytes \t Encoded size: 1992 bytes\n",
            "Original size: 381 bytes \t Encoded size: 180 bytes\n",
            "Original size: 68985 bytes \t Encoded size: 9388 bytes\n",
            "Original size: 1425 bytes \t Encoded size: 396 bytes\n",
            "Original size: 99501 bytes \t Encoded size: 14004 bytes\n",
            "Original size: 321 bytes \t Encoded size: 168 bytes\n",
            "Original size: 1413 bytes \t Encoded size: 424 bytes\n",
            "Number of labels: 19\n",
            "Original size: 218649 bytes \t Encoded size: 36104 bytes\n",
            "Original size: 215793 bytes \t Encoded size: 40832 bytes\n",
            "Original size: 18033 bytes \t Encoded size: 3928 bytes\n",
            "Original size: 13341 bytes \t Encoded size: 2460 bytes\n",
            "Original size: 1067397 bytes \t Encoded size: 99064 bytes\n",
            "Original size: 655341 bytes \t Encoded size: 84948 bytes\n",
            "Original size: 258813 bytes \t Encoded size: 52716 bytes\n",
            "Original size: 562653 bytes \t Encoded size: 70196 bytes\n",
            "Original size: 60825 bytes \t Encoded size: 11984 bytes\n",
            "Original size: 3717 bytes \t Encoded size: 788 bytes\n",
            "Original size: 1185 bytes \t Encoded size: 444 bytes\n",
            "Original size: 34605 bytes \t Encoded size: 7508 bytes\n",
            "Original size: 1197 bytes \t Encoded size: 388 bytes\n",
            "Original size: 88749 bytes \t Encoded size: 11600 bytes\n",
            "Original size: 7857 bytes \t Encoded size: 1260 bytes\n",
            "Original size: 259197 bytes \t Encoded size: 18996 bytes\n",
            "Original size: 549 bytes \t Encoded size: 276 bytes\n",
            "Original size: 1101 bytes \t Encoded size: 440 bytes\n",
            "Original size: 6753 bytes \t Encoded size: 1212 bytes\n",
            "Number of labels: 13\n",
            "Original size: 913965 bytes \t Encoded size: 143408 bytes\n",
            "Original size: 11601 bytes \t Encoded size: 3468 bytes\n",
            "Original size: 10017 bytes \t Encoded size: 1876 bytes\n",
            "Original size: 1195797 bytes \t Encoded size: 113668 bytes\n",
            "Original size: 638457 bytes \t Encoded size: 78964 bytes\n",
            "Original size: 120093 bytes \t Encoded size: 24428 bytes\n",
            "Original size: 562653 bytes \t Encoded size: 70192 bytes\n",
            "Original size: 7053 bytes \t Encoded size: 1680 bytes\n",
            "Original size: 1593 bytes \t Encoded size: 576 bytes\n",
            "Original size: 21537 bytes \t Encoded size: 4108 bytes\n",
            "Original size: 129 bytes \t Encoded size: 112 bytes\n",
            "Original size: 14409 bytes \t Encoded size: 3068 bytes\n",
            "Original size: 57 bytes \t Encoded size: 88 bytes\n",
            "Number of labels: 19\n",
            "Original size: 50589 bytes \t Encoded size: 10920 bytes\n",
            "Original size: 114933 bytes \t Encoded size: 24952 bytes\n",
            "Original size: 59421 bytes \t Encoded size: 10416 bytes\n",
            "Original size: 20553 bytes \t Encoded size: 3676 bytes\n",
            "Original size: 1263057 bytes \t Encoded size: 104588 bytes\n",
            "Original size: 377265 bytes \t Encoded size: 46380 bytes\n",
            "Original size: 283317 bytes \t Encoded size: 55500 bytes\n",
            "Original size: 562641 bytes \t Encoded size: 70208 bytes\n",
            "Original size: 18621 bytes \t Encoded size: 4216 bytes\n",
            "Original size: 345 bytes \t Encoded size: 156 bytes\n",
            "Original size: 207045 bytes \t Encoded size: 23564 bytes\n",
            "Original size: 2301 bytes \t Encoded size: 620 bytes\n",
            "Original size: 9033 bytes \t Encoded size: 2080 bytes\n",
            "Original size: 225 bytes \t Encoded size: 140 bytes\n",
            "Original size: 78033 bytes \t Encoded size: 10524 bytes\n",
            "Original size: 1905 bytes \t Encoded size: 584 bytes\n",
            "Original size: 228957 bytes \t Encoded size: 28712 bytes\n",
            "Original size: 1161 bytes \t Encoded size: 376 bytes\n",
            "Original size: 1629 bytes \t Encoded size: 512 bytes\n",
            "Number of labels: 21\n",
            "Original size: 33609 bytes \t Encoded size: 7564 bytes\n",
            "Original size: 149097 bytes \t Encoded size: 27000 bytes\n",
            "Original size: 16941 bytes \t Encoded size: 3444 bytes\n",
            "Original size: 13941 bytes \t Encoded size: 2728 bytes\n",
            "Original size: 1115253 bytes \t Encoded size: 92072 bytes\n",
            "Original size: 596685 bytes \t Encoded size: 64916 bytes\n",
            "Original size: 86193 bytes \t Encoded size: 18756 bytes\n",
            "Original size: 562653 bytes \t Encoded size: 70192 bytes\n",
            "Original size: 270921 bytes \t Encoded size: 34656 bytes\n",
            "Original size: 14865 bytes \t Encoded size: 3516 bytes\n",
            "Original size: 81 bytes \t Encoded size: 100 bytes\n",
            "Original size: 16173 bytes \t Encoded size: 2480 bytes\n",
            "Original size: 13077 bytes \t Encoded size: 1820 bytes\n",
            "Original size: 2229 bytes \t Encoded size: 684 bytes\n",
            "Original size: 24225 bytes \t Encoded size: 4280 bytes\n",
            "Original size: 321 bytes \t Encoded size: 172 bytes\n",
            "Original size: 273381 bytes \t Encoded size: 26816 bytes\n",
            "Original size: 1797 bytes \t Encoded size: 680 bytes\n",
            "Original size: 6009 bytes \t Encoded size: 1580 bytes\n",
            "Original size: 153 bytes \t Encoded size: 124 bytes\n",
            "Original size: 153 bytes \t Encoded size: 124 bytes\n",
            "Number of labels: 19\n",
            "Original size: 69453 bytes \t Encoded size: 14552 bytes\n",
            "Original size: 146013 bytes \t Encoded size: 30024 bytes\n",
            "Original size: 34017 bytes \t Encoded size: 6412 bytes\n",
            "Original size: 225 bytes \t Encoded size: 148 bytes\n",
            "Original size: 37245 bytes \t Encoded size: 4388 bytes\n",
            "Original size: 1442613 bytes \t Encoded size: 130652 bytes\n",
            "Original size: 505941 bytes \t Encoded size: 62628 bytes\n",
            "Original size: 76965 bytes \t Encoded size: 14920 bytes\n",
            "Original size: 562653 bytes \t Encoded size: 70168 bytes\n",
            "Original size: 31941 bytes \t Encoded size: 6932 bytes\n",
            "Original size: 321 bytes \t Encoded size: 192 bytes\n",
            "Original size: 18297 bytes \t Encoded size: 4216 bytes\n",
            "Original size: 2265 bytes \t Encoded size: 540 bytes\n",
            "Original size: 67377 bytes \t Encoded size: 11548 bytes\n",
            "Original size: 2073 bytes \t Encoded size: 592 bytes\n",
            "Original size: 52509 bytes \t Encoded size: 7572 bytes\n",
            "Original size: 41121 bytes \t Encoded size: 6940 bytes\n",
            "Original size: 6141 bytes \t Encoded size: 1792 bytes\n",
            "Original size: 2637 bytes \t Encoded size: 648 bytes\n",
            "Number of labels: 16\n",
            "Original size: 57549 bytes \t Encoded size: 12216 bytes\n",
            "Original size: 289569 bytes \t Encoded size: 54448 bytes\n",
            "Original size: 22401 bytes \t Encoded size: 5004 bytes\n",
            "Original size: 24585 bytes \t Encoded size: 3464 bytes\n",
            "Original size: 1290273 bytes \t Encoded size: 131508 bytes\n",
            "Original size: 530805 bytes \t Encoded size: 65580 bytes\n",
            "Original size: 263877 bytes \t Encoded size: 46212 bytes\n",
            "Original size: 562653 bytes \t Encoded size: 70180 bytes\n",
            "Original size: 69093 bytes \t Encoded size: 11540 bytes\n",
            "Original size: 501 bytes \t Encoded size: 240 bytes\n",
            "Original size: 2481 bytes \t Encoded size: 592 bytes\n",
            "Original size: 32061 bytes \t Encoded size: 4300 bytes\n",
            "Original size: 17973 bytes \t Encoded size: 4288 bytes\n",
            "Original size: 63873 bytes \t Encoded size: 9504 bytes\n",
            "Original size: 128181 bytes \t Encoded size: 24780 bytes\n",
            "Original size: 3741 bytes \t Encoded size: 1240 bytes\n",
            "Number of labels: 20\n",
            "Original size: 318993 bytes \t Encoded size: 57924 bytes\n",
            "Original size: 81561 bytes \t Encoded size: 18484 bytes\n",
            "Original size: 108213 bytes \t Encoded size: 15684 bytes\n",
            "Original size: 26133 bytes \t Encoded size: 4300 bytes\n",
            "Original size: 1376925 bytes \t Encoded size: 135036 bytes\n",
            "Original size: 509241 bytes \t Encoded size: 60676 bytes\n",
            "Original size: 208209 bytes \t Encoded size: 41788 bytes\n",
            "Original size: 562653 bytes \t Encoded size: 70216 bytes\n",
            "Original size: 33777 bytes \t Encoded size: 6748 bytes\n",
            "Original size: 45 bytes \t Encoded size: 80 bytes\n",
            "Original size: 45 bytes \t Encoded size: 80 bytes\n",
            "Original size: 381 bytes \t Encoded size: 204 bytes\n",
            "Original size: 2277 bytes \t Encoded size: 540 bytes\n",
            "Original size: 5733 bytes \t Encoded size: 1572 bytes\n",
            "Original size: 8025 bytes \t Encoded size: 2056 bytes\n",
            "Original size: 58953 bytes \t Encoded size: 9692 bytes\n",
            "Original size: 2073 bytes \t Encoded size: 508 bytes\n",
            "Original size: 79437 bytes \t Encoded size: 13772 bytes\n",
            "Original size: 237 bytes \t Encoded size: 152 bytes\n",
            "Original size: 1689 bytes \t Encoded size: 436 bytes\n",
            "Number of labels: 20\n",
            "Original size: 12225 bytes \t Encoded size: 2848 bytes\n",
            "Original size: 205677 bytes \t Encoded size: 39108 bytes\n",
            "Original size: 2589 bytes \t Encoded size: 708 bytes\n",
            "Original size: 3621 bytes \t Encoded size: 1208 bytes\n",
            "Original size: 24669 bytes \t Encoded size: 3516 bytes\n",
            "Original size: 1266489 bytes \t Encoded size: 121004 bytes\n",
            "Original size: 498585 bytes \t Encoded size: 60600 bytes\n",
            "Original size: 232653 bytes \t Encoded size: 34832 bytes\n",
            "Original size: 562653 bytes \t Encoded size: 70184 bytes\n",
            "Original size: 101493 bytes \t Encoded size: 14924 bytes\n",
            "Original size: 45 bytes \t Encoded size: 80 bytes\n",
            "Original size: 2529 bytes \t Encoded size: 552 bytes\n",
            "Original size: 24897 bytes \t Encoded size: 4292 bytes\n",
            "Original size: 25581 bytes \t Encoded size: 4720 bytes\n",
            "Original size: 52401 bytes \t Encoded size: 8456 bytes\n",
            "Original size: 6429 bytes \t Encoded size: 1248 bytes\n",
            "Original size: 264777 bytes \t Encoded size: 34972 bytes\n",
            "Original size: 20997 bytes \t Encoded size: 3484 bytes\n",
            "Original size: 4473 bytes \t Encoded size: 884 bytes\n",
            "Original size: 2673 bytes \t Encoded size: 648 bytes\n",
            "Number of labels: 19\n",
            "Original size: 6537 bytes \t Encoded size: 1480 bytes\n",
            "Original size: 191097 bytes \t Encoded size: 34456 bytes\n",
            "Original size: 71001 bytes \t Encoded size: 10536 bytes\n",
            "Original size: 8205 bytes \t Encoded size: 1548 bytes\n",
            "Original size: 1173513 bytes \t Encoded size: 98760 bytes\n",
            "Original size: 665457 bytes \t Encoded size: 78704 bytes\n",
            "Original size: 103629 bytes \t Encoded size: 17876 bytes\n",
            "Original size: 562653 bytes \t Encoded size: 70172 bytes\n",
            "Original size: 28353 bytes \t Encoded size: 5048 bytes\n",
            "Original size: 69 bytes \t Encoded size: 92 bytes\n",
            "Original size: 2253 bytes \t Encoded size: 664 bytes\n",
            "Original size: 2169 bytes \t Encoded size: 552 bytes\n",
            "Original size: 13917 bytes \t Encoded size: 2420 bytes\n",
            "Original size: 2565 bytes \t Encoded size: 608 bytes\n",
            "Original size: 25641 bytes \t Encoded size: 4156 bytes\n",
            "Original size: 477 bytes \t Encoded size: 188 bytes\n",
            "Original size: 64821 bytes \t Encoded size: 10920 bytes\n",
            "Original size: 1665 bytes \t Encoded size: 412 bytes\n",
            "Original size: 141 bytes \t Encoded size: 116 bytes\n",
            "Number of labels: 20\n",
            "Original size: 29373 bytes \t Encoded size: 6132 bytes\n",
            "Original size: 254409 bytes \t Encoded size: 43148 bytes\n",
            "Original size: 68805 bytes \t Encoded size: 10276 bytes\n",
            "Original size: 813 bytes \t Encoded size: 292 bytes\n",
            "Original size: 3801 bytes \t Encoded size: 996 bytes\n",
            "Original size: 1135893 bytes \t Encoded size: 94688 bytes\n",
            "Original size: 637305 bytes \t Encoded size: 74444 bytes\n",
            "Original size: 29457 bytes \t Encoded size: 6732 bytes\n",
            "Original size: 562653 bytes \t Encoded size: 70196 bytes\n",
            "Original size: 62577 bytes \t Encoded size: 10652 bytes\n",
            "Original size: 513 bytes \t Encoded size: 248 bytes\n",
            "Original size: 273 bytes \t Encoded size: 168 bytes\n",
            "Original size: 309 bytes \t Encoded size: 172 bytes\n",
            "Original size: 9837 bytes \t Encoded size: 1716 bytes\n",
            "Original size: 45 bytes \t Encoded size: 80 bytes\n",
            "Original size: 35085 bytes \t Encoded size: 5388 bytes\n",
            "Original size: 357 bytes \t Encoded size: 184 bytes\n",
            "Original size: 132213 bytes \t Encoded size: 15432 bytes\n",
            "Original size: 2001 bytes \t Encoded size: 732 bytes\n",
            "Original size: 405 bytes \t Encoded size: 196 bytes\n",
            "Number of labels: 21\n",
            "Original size: 18273 bytes \t Encoded size: 4420 bytes\n",
            "Original size: 329757 bytes \t Encoded size: 61792 bytes\n",
            "Original size: 15369 bytes \t Encoded size: 3148 bytes\n",
            "Original size: 18753 bytes \t Encoded size: 3056 bytes\n",
            "Original size: 1174101 bytes \t Encoded size: 105092 bytes\n",
            "Original size: 679161 bytes \t Encoded size: 84696 bytes\n",
            "Original size: 50805 bytes \t Encoded size: 11588 bytes\n",
            "Original size: 562653 bytes \t Encoded size: 70196 bytes\n",
            "Original size: 4401 bytes \t Encoded size: 1220 bytes\n",
            "Original size: 55005 bytes \t Encoded size: 10620 bytes\n",
            "Original size: 2145 bytes \t Encoded size: 504 bytes\n",
            "Original size: 107925 bytes \t Encoded size: 13900 bytes\n",
            "Original size: 657 bytes \t Encoded size: 248 bytes\n",
            "Original size: 309 bytes \t Encoded size: 168 bytes\n",
            "Original size: 19509 bytes \t Encoded size: 3644 bytes\n",
            "Original size: 18081 bytes \t Encoded size: 2796 bytes\n",
            "Original size: 189369 bytes \t Encoded size: 22944 bytes\n",
            "Original size: 2745 bytes \t Encoded size: 768 bytes\n",
            "Original size: 3825 bytes \t Encoded size: 1024 bytes\n",
            "Original size: 129 bytes \t Encoded size: 116 bytes\n",
            "Original size: 19161 bytes \t Encoded size: 2220 bytes\n",
            "Number of labels: 20\n",
            "Original size: 43605 bytes \t Encoded size: 9980 bytes\n",
            "Original size: 122025 bytes \t Encoded size: 25756 bytes\n",
            "Original size: 5769 bytes \t Encoded size: 1492 bytes\n",
            "Original size: 177 bytes \t Encoded size: 132 bytes\n",
            "Original size: 14481 bytes \t Encoded size: 2924 bytes\n",
            "Original size: 1551705 bytes \t Encoded size: 138172 bytes\n",
            "Original size: 345225 bytes \t Encoded size: 52168 bytes\n",
            "Original size: 178797 bytes \t Encoded size: 38112 bytes\n",
            "Original size: 562653 bytes \t Encoded size: 70176 bytes\n",
            "Original size: 47349 bytes \t Encoded size: 6832 bytes\n",
            "Original size: 18933 bytes \t Encoded size: 4528 bytes\n",
            "Original size: 957 bytes \t Encoded size: 372 bytes\n",
            "Original size: 2757 bytes \t Encoded size: 640 bytes\n",
            "Original size: 39789 bytes \t Encoded size: 5520 bytes\n",
            "Original size: 3825 bytes \t Encoded size: 852 bytes\n",
            "Original size: 64041 bytes \t Encoded size: 9664 bytes\n",
            "Original size: 117 bytes \t Encoded size: 108 bytes\n",
            "Original size: 239697 bytes \t Encoded size: 33608 bytes\n",
            "Original size: 1725 bytes \t Encoded size: 532 bytes\n",
            "Original size: 57 bytes \t Encoded size: 88 bytes\n",
            "Number of labels: 18\n",
            "Original size: 175173 bytes \t Encoded size: 34020 bytes\n",
            "Original size: 42981 bytes \t Encoded size: 10092 bytes\n",
            "Original size: 108609 bytes \t Encoded size: 17904 bytes\n",
            "Original size: 14097 bytes \t Encoded size: 2548 bytes\n",
            "Original size: 1430565 bytes \t Encoded size: 127528 bytes\n",
            "Original size: 405981 bytes \t Encoded size: 56676 bytes\n",
            "Original size: 127977 bytes \t Encoded size: 29464 bytes\n",
            "Original size: 562653 bytes \t Encoded size: 70176 bytes\n",
            "Original size: 22521 bytes \t Encoded size: 4712 bytes\n",
            "Original size: 1053 bytes \t Encoded size: 312 bytes\n",
            "Original size: 44985 bytes \t Encoded size: 7628 bytes\n",
            "Original size: 2661 bytes \t Encoded size: 620 bytes\n",
            "Original size: 5529 bytes \t Encoded size: 1496 bytes\n",
            "Original size: 5733 bytes \t Encoded size: 1692 bytes\n",
            "Original size: 59865 bytes \t Encoded size: 8596 bytes\n",
            "Original size: 261 bytes \t Encoded size: 160 bytes\n",
            "Original size: 118989 bytes \t Encoded size: 16704 bytes\n",
            "Original size: 93 bytes \t Encoded size: 100 bytes\n",
            "Number of labels: 15\n",
            "Original size: 45453 bytes \t Encoded size: 9360 bytes\n",
            "Original size: 224565 bytes \t Encoded size: 43936 bytes\n",
            "Original size: 11277 bytes \t Encoded size: 2744 bytes\n",
            "Original size: 10137 bytes \t Encoded size: 2168 bytes\n",
            "Original size: 1196253 bytes \t Encoded size: 118480 bytes\n",
            "Original size: 628029 bytes \t Encoded size: 84712 bytes\n",
            "Original size: 145065 bytes \t Encoded size: 22236 bytes\n",
            "Original size: 562653 bytes \t Encoded size: 70184 bytes\n",
            "Original size: 189921 bytes \t Encoded size: 31768 bytes\n",
            "Original size: 3705 bytes \t Encoded size: 1112 bytes\n",
            "Original size: 3417 bytes \t Encoded size: 744 bytes\n",
            "Original size: 10497 bytes \t Encoded size: 2156 bytes\n",
            "Original size: 11085 bytes \t Encoded size: 2696 bytes\n",
            "Original size: 106341 bytes \t Encoded size: 14508 bytes\n",
            "Original size: 64053 bytes \t Encoded size: 11476 bytes\n",
            "Number of labels: 22\n",
            "Original size: 75885 bytes \t Encoded size: 15300 bytes\n",
            "Original size: 153189 bytes \t Encoded size: 31472 bytes\n",
            "Original size: 26253 bytes \t Encoded size: 4616 bytes\n",
            "Original size: 333 bytes \t Encoded size: 184 bytes\n",
            "Original size: 21765 bytes \t Encoded size: 3592 bytes\n",
            "Original size: 1373313 bytes \t Encoded size: 136724 bytes\n",
            "Original size: 440337 bytes \t Encoded size: 56780 bytes\n",
            "Original size: 94761 bytes \t Encoded size: 17648 bytes\n",
            "Original size: 562653 bytes \t Encoded size: 70188 bytes\n",
            "Original size: 20145 bytes \t Encoded size: 4540 bytes\n",
            "Original size: 477 bytes \t Encoded size: 236 bytes\n",
            "Original size: 57645 bytes \t Encoded size: 11548 bytes\n",
            "Original size: 2517 bytes \t Encoded size: 508 bytes\n",
            "Original size: 53193 bytes \t Encoded size: 8908 bytes\n",
            "Original size: 1221 bytes \t Encoded size: 380 bytes\n",
            "Original size: 1173 bytes \t Encoded size: 448 bytes\n",
            "Original size: 41613 bytes \t Encoded size: 5816 bytes\n",
            "Original size: 2445 bytes \t Encoded size: 620 bytes\n",
            "Original size: 243537 bytes \t Encoded size: 22516 bytes\n",
            "Original size: 3333 bytes \t Encoded size: 1088 bytes\n",
            "Original size: 909 bytes \t Encoded size: 384 bytes\n",
            "Original size: 933 bytes \t Encoded size: 316 bytes\n",
            "Number of labels: 18\n",
            "Original size: 1425 bytes \t Encoded size: 464 bytes\n",
            "Original size: 424293 bytes \t Encoded size: 73440 bytes\n",
            "Original size: 65541 bytes \t Encoded size: 10800 bytes\n",
            "Original size: 19533 bytes \t Encoded size: 3032 bytes\n",
            "Original size: 1204413 bytes \t Encoded size: 104048 bytes\n",
            "Original size: 647169 bytes \t Encoded size: 75572 bytes\n",
            "Original size: 214689 bytes \t Encoded size: 41268 bytes\n",
            "Original size: 562653 bytes \t Encoded size: 70180 bytes\n",
            "Original size: 7353 bytes \t Encoded size: 1700 bytes\n",
            "Original size: 3729 bytes \t Encoded size: 780 bytes\n",
            "Original size: 5769 bytes \t Encoded size: 1352 bytes\n",
            "Original size: 1713 bytes \t Encoded size: 440 bytes\n",
            "Original size: 801 bytes \t Encoded size: 296 bytes\n",
            "Original size: 1041 bytes \t Encoded size: 332 bytes\n",
            "Original size: 89097 bytes \t Encoded size: 10460 bytes\n",
            "Original size: 609 bytes \t Encoded size: 288 bytes\n",
            "Original size: 27273 bytes \t Encoded size: 4960 bytes\n",
            "Original size: 393 bytes \t Encoded size: 184 bytes\n",
            "Number of labels: 18\n",
            "Original size: 22401 bytes \t Encoded size: 5244 bytes\n",
            "Original size: 138789 bytes \t Encoded size: 25784 bytes\n",
            "Original size: 28689 bytes \t Encoded size: 5492 bytes\n",
            "Original size: 11457 bytes \t Encoded size: 2252 bytes\n",
            "Original size: 1414425 bytes \t Encoded size: 122376 bytes\n",
            "Original size: 361017 bytes \t Encoded size: 50000 bytes\n",
            "Original size: 120177 bytes \t Encoded size: 17812 bytes\n",
            "Original size: 562653 bytes \t Encoded size: 70180 bytes\n",
            "Original size: 60933 bytes \t Encoded size: 10096 bytes\n",
            "Original size: 18597 bytes \t Encoded size: 3292 bytes\n",
            "Original size: 3357 bytes \t Encoded size: 740 bytes\n",
            "Original size: 36741 bytes \t Encoded size: 4824 bytes\n",
            "Original size: 1329 bytes \t Encoded size: 432 bytes\n",
            "Original size: 32493 bytes \t Encoded size: 5464 bytes\n",
            "Original size: 1005 bytes \t Encoded size: 304 bytes\n",
            "Original size: 288465 bytes \t Encoded size: 27840 bytes\n",
            "Original size: 501 bytes \t Encoded size: 244 bytes\n",
            "Original size: 609 bytes \t Encoded size: 236 bytes\n",
            "Number of labels: 20\n",
            "Original size: 76329 bytes \t Encoded size: 14560 bytes\n",
            "Original size: 56817 bytes \t Encoded size: 12116 bytes\n",
            "Original size: 10125 bytes \t Encoded size: 2336 bytes\n",
            "Original size: 14325 bytes \t Encoded size: 2700 bytes\n",
            "Original size: 1138413 bytes \t Encoded size: 94668 bytes\n",
            "Original size: 629289 bytes \t Encoded size: 73900 bytes\n",
            "Original size: 224025 bytes \t Encoded size: 46620 bytes\n",
            "Original size: 562653 bytes \t Encoded size: 70184 bytes\n",
            "Original size: 18921 bytes \t Encoded size: 3756 bytes\n",
            "Original size: 573 bytes \t Encoded size: 276 bytes\n",
            "Original size: 268041 bytes \t Encoded size: 41088 bytes\n",
            "Original size: 57 bytes \t Encoded size: 88 bytes\n",
            "Original size: 9117 bytes \t Encoded size: 2028 bytes\n",
            "Original size: 1521 bytes \t Encoded size: 584 bytes\n",
            "Original size: 5373 bytes \t Encoded size: 1352 bytes\n",
            "Original size: 3501 bytes \t Encoded size: 808 bytes\n",
            "Original size: 56313 bytes \t Encoded size: 5944 bytes\n",
            "Original size: 126621 bytes \t Encoded size: 32256 bytes\n",
            "Original size: 609 bytes \t Encoded size: 288 bytes\n",
            "Original size: 2349 bytes \t Encoded size: 648 bytes\n",
            "Number of labels: 18\n",
            "Original size: 276285 bytes \t Encoded size: 55956 bytes\n",
            "Original size: 5133 bytes \t Encoded size: 1668 bytes\n",
            "Original size: 4917 bytes \t Encoded size: 1124 bytes\n",
            "Original size: 1117605 bytes \t Encoded size: 95080 bytes\n",
            "Original size: 387117 bytes \t Encoded size: 43356 bytes\n",
            "Original size: 247053 bytes \t Encoded size: 52500 bytes\n",
            "Original size: 562653 bytes \t Encoded size: 70188 bytes\n",
            "Original size: 453 bytes \t Encoded size: 208 bytes\n",
            "Original size: 160017 bytes \t Encoded size: 28868 bytes\n",
            "Original size: 356373 bytes \t Encoded size: 43444 bytes\n",
            "Original size: 129 bytes \t Encoded size: 112 bytes\n",
            "Original size: 513 bytes \t Encoded size: 236 bytes\n",
            "Original size: 2589 bytes \t Encoded size: 904 bytes\n",
            "Original size: 1209 bytes \t Encoded size: 432 bytes\n",
            "Original size: 345 bytes \t Encoded size: 168 bytes\n",
            "Original size: 256485 bytes \t Encoded size: 29588 bytes\n",
            "Original size: 45 bytes \t Encoded size: 80 bytes\n",
            "Original size: 117 bytes \t Encoded size: 112 bytes\n",
            "Number of labels: 20\n",
            "Original size: 106761 bytes \t Encoded size: 21340 bytes\n",
            "Original size: 165729 bytes \t Encoded size: 35024 bytes\n",
            "Original size: 47109 bytes \t Encoded size: 7864 bytes\n",
            "Original size: 17841 bytes \t Encoded size: 3260 bytes\n",
            "Original size: 1189125 bytes \t Encoded size: 102172 bytes\n",
            "Original size: 660633 bytes \t Encoded size: 71792 bytes\n",
            "Original size: 88389 bytes \t Encoded size: 19016 bytes\n",
            "Original size: 562653 bytes \t Encoded size: 70196 bytes\n",
            "Original size: 114009 bytes \t Encoded size: 18216 bytes\n",
            "Original size: 46245 bytes \t Encoded size: 11096 bytes\n",
            "Original size: 45 bytes \t Encoded size: 80 bytes\n",
            "Original size: 2961 bytes \t Encoded size: 904 bytes\n",
            "Original size: 3981 bytes \t Encoded size: 860 bytes\n",
            "Original size: 14169 bytes \t Encoded size: 3936 bytes\n",
            "Original size: 79713 bytes \t Encoded size: 11060 bytes\n",
            "Original size: 5649 bytes \t Encoded size: 984 bytes\n",
            "Original size: 15657 bytes \t Encoded size: 3036 bytes\n",
            "Original size: 13641 bytes \t Encoded size: 3108 bytes\n",
            "Original size: 57 bytes \t Encoded size: 88 bytes\n",
            "Original size: 2325 bytes \t Encoded size: 532 bytes\n",
            "Number of labels: 14\n",
            "Original size: 588669 bytes \t Encoded size: 98660 bytes\n",
            "Original size: 214305 bytes \t Encoded size: 36248 bytes\n",
            "Original size: 10149 bytes \t Encoded size: 2048 bytes\n",
            "Original size: 1241901 bytes \t Encoded size: 132840 bytes\n",
            "Original size: 594969 bytes \t Encoded size: 80096 bytes\n",
            "Original size: 79125 bytes \t Encoded size: 16476 bytes\n",
            "Original size: 562653 bytes \t Encoded size: 70188 bytes\n",
            "Original size: 1557 bytes \t Encoded size: 472 bytes\n",
            "Original size: 76785 bytes \t Encoded size: 14048 bytes\n",
            "Original size: 93 bytes \t Encoded size: 104 bytes\n",
            "Original size: 32481 bytes \t Encoded size: 5788 bytes\n",
            "Original size: 369 bytes \t Encoded size: 180 bytes\n",
            "Original size: 27933 bytes \t Encoded size: 5044 bytes\n",
            "Original size: 225 bytes \t Encoded size: 136 bytes\n",
            "Number of labels: 17\n",
            "Original size: 148269 bytes \t Encoded size: 29044 bytes\n",
            "Original size: 345057 bytes \t Encoded size: 62808 bytes\n",
            "Original size: 51825 bytes \t Encoded size: 8560 bytes\n",
            "Original size: 16125 bytes \t Encoded size: 2608 bytes\n",
            "Original size: 1120461 bytes \t Encoded size: 96824 bytes\n",
            "Original size: 642609 bytes \t Encoded size: 85524 bytes\n",
            "Original size: 276069 bytes \t Encoded size: 53908 bytes\n",
            "Original size: 562653 bytes \t Encoded size: 70196 bytes\n",
            "Original size: 39453 bytes \t Encoded size: 7892 bytes\n",
            "Original size: 117 bytes \t Encoded size: 112 bytes\n",
            "Original size: 933 bytes \t Encoded size: 316 bytes\n",
            "Original size: 12537 bytes \t Encoded size: 2068 bytes\n",
            "Original size: 39609 bytes \t Encoded size: 5344 bytes\n",
            "Original size: 5349 bytes \t Encoded size: 1016 bytes\n",
            "Original size: 193509 bytes \t Encoded size: 23248 bytes\n",
            "Original size: 165 bytes \t Encoded size: 128 bytes\n",
            "Original size: 4437 bytes \t Encoded size: 924 bytes\n",
            "Number of labels: 21\n",
            "Original size: 142029 bytes \t Encoded size: 26712 bytes\n",
            "Original size: 49029 bytes \t Encoded size: 10860 bytes\n",
            "Original size: 4413 bytes \t Encoded size: 1216 bytes\n",
            "Original size: 13005 bytes \t Encoded size: 2540 bytes\n",
            "Original size: 1188249 bytes \t Encoded size: 106476 bytes\n",
            "Original size: 633573 bytes \t Encoded size: 73076 bytes\n",
            "Original size: 224961 bytes \t Encoded size: 45844 bytes\n",
            "Original size: 562653 bytes \t Encoded size: 70176 bytes\n",
            "Original size: 6429 bytes \t Encoded size: 1644 bytes\n",
            "Original size: 141 bytes \t Encoded size: 116 bytes\n",
            "Original size: 326457 bytes \t Encoded size: 56596 bytes\n",
            "Original size: 237 bytes \t Encoded size: 152 bytes\n",
            "Original size: 1233 bytes \t Encoded size: 372 bytes\n",
            "Original size: 1845 bytes \t Encoded size: 564 bytes\n",
            "Original size: 213 bytes \t Encoded size: 128 bytes\n",
            "Original size: 1317 bytes \t Encoded size: 472 bytes\n",
            "Original size: 9141 bytes \t Encoded size: 2232 bytes\n",
            "Original size: 573 bytes \t Encoded size: 244 bytes\n",
            "Original size: 64953 bytes \t Encoded size: 10860 bytes\n",
            "Original size: 321 bytes \t Encoded size: 168 bytes\n",
            "Original size: 117 bytes \t Encoded size: 108 bytes\n",
            "Number of labels: 20\n",
            "Original size: 561 bytes \t Encoded size: 228 bytes\n",
            "Original size: 520869 bytes \t Encoded size: 85884 bytes\n",
            "Original size: 52737 bytes \t Encoded size: 4628 bytes\n",
            "Original size: 8133 bytes \t Encoded size: 1668 bytes\n",
            "Original size: 1116081 bytes \t Encoded size: 90148 bytes\n",
            "Original size: 680493 bytes \t Encoded size: 86056 bytes\n",
            "Original size: 69 bytes \t Encoded size: 92 bytes\n",
            "Original size: 562653 bytes \t Encoded size: 70192 bytes\n",
            "Original size: 43281 bytes \t Encoded size: 8132 bytes\n",
            "Original size: 189 bytes \t Encoded size: 136 bytes\n",
            "Original size: 333 bytes \t Encoded size: 168 bytes\n",
            "Original size: 40929 bytes \t Encoded size: 6088 bytes\n",
            "Original size: 12777 bytes \t Encoded size: 2320 bytes\n",
            "Original size: 40593 bytes \t Encoded size: 6360 bytes\n",
            "Original size: 1893 bytes \t Encoded size: 436 bytes\n",
            "Original size: 28233 bytes \t Encoded size: 4856 bytes\n",
            "Original size: 1377 bytes \t Encoded size: 380 bytes\n",
            "Original size: 285873 bytes \t Encoded size: 54236 bytes\n",
            "Original size: 429 bytes \t Encoded size: 192 bytes\n",
            "Original size: 1701 bytes \t Encoded size: 520 bytes\n",
            "Number of labels: 20\n",
            "Original size: 21837 bytes \t Encoded size: 5348 bytes\n",
            "Original size: 239265 bytes \t Encoded size: 46924 bytes\n",
            "Original size: 7161 bytes \t Encoded size: 1768 bytes\n",
            "Original size: 32817 bytes \t Encoded size: 3756 bytes\n",
            "Original size: 1183281 bytes \t Encoded size: 104112 bytes\n",
            "Original size: 657429 bytes \t Encoded size: 73908 bytes\n",
            "Original size: 71397 bytes \t Encoded size: 16568 bytes\n",
            "Original size: 562653 bytes \t Encoded size: 70196 bytes\n",
            "Original size: 5409 bytes \t Encoded size: 1480 bytes\n",
            "Original size: 57 bytes \t Encoded size: 88 bytes\n",
            "Original size: 227505 bytes \t Encoded size: 35104 bytes\n",
            "Original size: 2577 bytes \t Encoded size: 540 bytes\n",
            "Original size: 59181 bytes \t Encoded size: 8692 bytes\n",
            "Original size: 405 bytes \t Encoded size: 192 bytes\n",
            "Original size: 2577 bytes \t Encoded size: 928 bytes\n",
            "Original size: 14121 bytes \t Encoded size: 2840 bytes\n",
            "Original size: 2373 bytes \t Encoded size: 516 bytes\n",
            "Original size: 120141 bytes \t Encoded size: 17432 bytes\n",
            "Original size: 4197 bytes \t Encoded size: 1108 bytes\n",
            "Original size: 2253 bytes \t Encoded size: 472 bytes\n",
            "Number of labels: 19\n",
            "Original size: 378405 bytes \t Encoded size: 58128 bytes\n",
            "Original size: 20793 bytes \t Encoded size: 5580 bytes\n",
            "Original size: 73521 bytes \t Encoded size: 14464 bytes\n",
            "Original size: 10509 bytes \t Encoded size: 2136 bytes\n",
            "Original size: 1216353 bytes \t Encoded size: 113092 bytes\n",
            "Original size: 598293 bytes \t Encoded size: 71676 bytes\n",
            "Original size: 119685 bytes \t Encoded size: 26820 bytes\n",
            "Original size: 562653 bytes \t Encoded size: 70196 bytes\n",
            "Original size: 5013 bytes \t Encoded size: 1440 bytes\n",
            "Original size: 873 bytes \t Encoded size: 276 bytes\n",
            "Original size: 37245 bytes \t Encoded size: 6876 bytes\n",
            "Original size: 2685 bytes \t Encoded size: 620 bytes\n",
            "Original size: 1461 bytes \t Encoded size: 528 bytes\n",
            "Original size: 2145 bytes \t Encoded size: 760 bytes\n",
            "Original size: 22305 bytes \t Encoded size: 4180 bytes\n",
            "Original size: 5733 bytes \t Encoded size: 1248 bytes\n",
            "Original size: 85689 bytes \t Encoded size: 13264 bytes\n",
            "Original size: 15273 bytes \t Encoded size: 3656 bytes\n",
            "Original size: 297 bytes \t Encoded size: 168 bytes\n",
            "Number of labels: 20\n",
            "Original size: 212301 bytes \t Encoded size: 37388 bytes\n",
            "Original size: 153897 bytes \t Encoded size: 33852 bytes\n",
            "Original size: 25989 bytes \t Encoded size: 5224 bytes\n",
            "Original size: 12573 bytes \t Encoded size: 2512 bytes\n",
            "Original size: 1202793 bytes \t Encoded size: 110004 bytes\n",
            "Original size: 594969 bytes \t Encoded size: 76208 bytes\n",
            "Original size: 198441 bytes \t Encoded size: 36496 bytes\n",
            "Original size: 562653 bytes \t Encoded size: 70184 bytes\n",
            "Original size: 65169 bytes \t Encoded size: 12968 bytes\n",
            "Original size: 93 bytes \t Encoded size: 100 bytes\n",
            "Original size: 2217 bytes \t Encoded size: 556 bytes\n",
            "Original size: 2769 bytes \t Encoded size: 736 bytes\n",
            "Original size: 8553 bytes \t Encoded size: 1936 bytes\n",
            "Original size: 201 bytes \t Encoded size: 132 bytes\n",
            "Original size: 30117 bytes \t Encoded size: 5852 bytes\n",
            "Original size: 345 bytes \t Encoded size: 184 bytes\n",
            "Original size: 205653 bytes \t Encoded size: 29576 bytes\n",
            "Original size: 5097 bytes \t Encoded size: 1584 bytes\n",
            "Original size: 225 bytes \t Encoded size: 148 bytes\n",
            "Original size: 4053 bytes \t Encoded size: 1124 bytes\n",
            "Number of labels: 18\n",
            "Original size: 260685 bytes \t Encoded size: 46884 bytes\n",
            "Original size: 91029 bytes \t Encoded size: 20384 bytes\n",
            "Original size: 117261 bytes \t Encoded size: 18120 bytes\n",
            "Original size: 22713 bytes \t Encoded size: 3768 bytes\n",
            "Original size: 1265337 bytes \t Encoded size: 116916 bytes\n",
            "Original size: 533421 bytes \t Encoded size: 62868 bytes\n",
            "Original size: 225525 bytes \t Encoded size: 45472 bytes\n",
            "Original size: 562653 bytes \t Encoded size: 70184 bytes\n",
            "Original size: 38553 bytes \t Encoded size: 7628 bytes\n",
            "Original size: 69 bytes \t Encoded size: 92 bytes\n",
            "Original size: 1785 bytes \t Encoded size: 436 bytes\n",
            "Original size: 7041 bytes \t Encoded size: 1880 bytes\n",
            "Original size: 8145 bytes \t Encoded size: 2048 bytes\n",
            "Original size: 45045 bytes \t Encoded size: 7440 bytes\n",
            "Original size: 225 bytes \t Encoded size: 132 bytes\n",
            "Original size: 213273 bytes \t Encoded size: 28388 bytes\n",
            "Original size: 393 bytes \t Encoded size: 216 bytes\n",
            "Original size: 129 bytes \t Encoded size: 116 bytes\n",
            "Number of labels: 19\n",
            "Original size: 25461 bytes \t Encoded size: 6212 bytes\n",
            "Original size: 281097 bytes \t Encoded size: 52884 bytes\n",
            "Original size: 12729 bytes \t Encoded size: 2796 bytes\n",
            "Original size: 19425 bytes \t Encoded size: 3568 bytes\n",
            "Original size: 1465665 bytes \t Encoded size: 128696 bytes\n",
            "Original size: 380109 bytes \t Encoded size: 55992 bytes\n",
            "Original size: 59145 bytes \t Encoded size: 13188 bytes\n",
            "Original size: 562653 bytes \t Encoded size: 70176 bytes\n",
            "Original size: 4977 bytes \t Encoded size: 1400 bytes\n",
            "Original size: 40653 bytes \t Encoded size: 7880 bytes\n",
            "Original size: 2913 bytes \t Encoded size: 692 bytes\n",
            "Original size: 46797 bytes \t Encoded size: 7516 bytes\n",
            "Original size: 945 bytes \t Encoded size: 312 bytes\n",
            "Original size: 39513 bytes \t Encoded size: 5516 bytes\n",
            "Original size: 7545 bytes \t Encoded size: 1468 bytes\n",
            "Original size: 279777 bytes \t Encoded size: 33532 bytes\n",
            "Original size: 537 bytes \t Encoded size: 264 bytes\n",
            "Original size: 5985 bytes \t Encoded size: 1540 bytes\n",
            "Original size: 5889 bytes \t Encoded size: 1232 bytes\n",
            "Number of labels: 14\n",
            "Original size: 839529 bytes \t Encoded size: 148228 bytes\n",
            "Original size: 1893 bytes \t Encoded size: 712 bytes\n",
            "Original size: 2049 bytes \t Encoded size: 512 bytes\n",
            "Original size: 1146045 bytes \t Encoded size: 93884 bytes\n",
            "Original size: 606897 bytes \t Encoded size: 70784 bytes\n",
            "Original size: 98637 bytes \t Encoded size: 19608 bytes\n",
            "Original size: 562653 bytes \t Encoded size: 70204 bytes\n",
            "Original size: 273 bytes \t Encoded size: 160 bytes\n",
            "Original size: 189 bytes \t Encoded size: 136 bytes\n",
            "Original size: 261 bytes \t Encoded size: 152 bytes\n",
            "Original size: 54081 bytes \t Encoded size: 7540 bytes\n",
            "Original size: 7881 bytes \t Encoded size: 1268 bytes\n",
            "Original size: 173877 bytes \t Encoded size: 20920 bytes\n",
            "Original size: 2337 bytes \t Encoded size: 484 bytes\n",
            "Number of labels: 20\n",
            "Original size: 52185 bytes \t Encoded size: 9096 bytes\n",
            "Original size: 209877 bytes \t Encoded size: 38760 bytes\n",
            "Original size: 42969 bytes \t Encoded size: 7704 bytes\n",
            "Original size: 6525 bytes \t Encoded size: 1380 bytes\n",
            "Original size: 1135641 bytes \t Encoded size: 103920 bytes\n",
            "Original size: 602841 bytes \t Encoded size: 79176 bytes\n",
            "Original size: 37869 bytes \t Encoded size: 8480 bytes\n",
            "Original size: 562653 bytes \t Encoded size: 70184 bytes\n",
            "Original size: 142497 bytes \t Encoded size: 22668 bytes\n",
            "Original size: 45 bytes \t Encoded size: 80 bytes\n",
            "Original size: 2253 bytes \t Encoded size: 820 bytes\n",
            "Original size: 1965 bytes \t Encoded size: 568 bytes\n",
            "Original size: 4665 bytes \t Encoded size: 988 bytes\n",
            "Original size: 2469 bytes \t Encoded size: 712 bytes\n",
            "Original size: 58689 bytes \t Encoded size: 8768 bytes\n",
            "Original size: 645 bytes \t Encoded size: 236 bytes\n",
            "Original size: 141189 bytes \t Encoded size: 18228 bytes\n",
            "Original size: 79041 bytes \t Encoded size: 14840 bytes\n",
            "Original size: 1761 bytes \t Encoded size: 664 bytes\n",
            "Original size: 9657 bytes \t Encoded size: 2216 bytes\n",
            "Number of labels: 21\n",
            "Original size: 454665 bytes \t Encoded size: 78400 bytes\n",
            "Original size: 31137 bytes \t Encoded size: 7660 bytes\n",
            "Original size: 90753 bytes \t Encoded size: 14276 bytes\n",
            "Original size: 17277 bytes \t Encoded size: 3236 bytes\n",
            "Original size: 1464861 bytes \t Encoded size: 133188 bytes\n",
            "Original size: 366021 bytes \t Encoded size: 51792 bytes\n",
            "Original size: 189777 bytes \t Encoded size: 35736 bytes\n",
            "Original size: 562653 bytes \t Encoded size: 70184 bytes\n",
            "Original size: 11517 bytes \t Encoded size: 2840 bytes\n",
            "Original size: 81 bytes \t Encoded size: 100 bytes\n",
            "Original size: 105 bytes \t Encoded size: 108 bytes\n",
            "Original size: 273 bytes \t Encoded size: 156 bytes\n",
            "Original size: 2265 bytes \t Encoded size: 560 bytes\n",
            "Original size: 2253 bytes \t Encoded size: 628 bytes\n",
            "Original size: 3657 bytes \t Encoded size: 988 bytes\n",
            "Original size: 55437 bytes \t Encoded size: 9312 bytes\n",
            "Original size: 1389 bytes \t Encoded size: 404 bytes\n",
            "Original size: 146325 bytes \t Encoded size: 15236 bytes\n",
            "Original size: 4713 bytes \t Encoded size: 1608 bytes\n",
            "Original size: 1353 bytes \t Encoded size: 404 bytes\n",
            "Original size: 345 bytes \t Encoded size: 164 bytes\n",
            "Number of labels: 20\n",
            "Original size: 271029 bytes \t Encoded size: 42732 bytes\n",
            "Original size: 37869 bytes \t Encoded size: 8836 bytes\n",
            "Original size: 7113 bytes \t Encoded size: 1704 bytes\n",
            "Original size: 33825 bytes \t Encoded size: 3936 bytes\n",
            "Original size: 1134717 bytes \t Encoded size: 99364 bytes\n",
            "Original size: 608265 bytes \t Encoded size: 70444 bytes\n",
            "Original size: 172233 bytes \t Encoded size: 36500 bytes\n",
            "Original size: 562653 bytes \t Encoded size: 70192 bytes\n",
            "Original size: 8181 bytes \t Encoded size: 2112 bytes\n",
            "Original size: 189 bytes \t Encoded size: 136 bytes\n",
            "Original size: 242637 bytes \t Encoded size: 40612 bytes\n",
            "Original size: 357 bytes \t Encoded size: 196 bytes\n",
            "Original size: 1689 bytes \t Encoded size: 404 bytes\n",
            "Original size: 2073 bytes \t Encoded size: 628 bytes\n",
            "Original size: 261 bytes \t Encoded size: 148 bytes\n",
            "Original size: 525 bytes \t Encoded size: 256 bytes\n",
            "Original size: 18765 bytes \t Encoded size: 3744 bytes\n",
            "Original size: 309 bytes \t Encoded size: 164 bytes\n",
            "Original size: 185433 bytes \t Encoded size: 21488 bytes\n",
            "Original size: 165 bytes \t Encoded size: 124 bytes\n",
            "Number of labels: 16\n",
            "Original size: 174273 bytes \t Encoded size: 32620 bytes\n",
            "Original size: 365025 bytes \t Encoded size: 67992 bytes\n",
            "Original size: 23325 bytes \t Encoded size: 4528 bytes\n",
            "Original size: 10137 bytes \t Encoded size: 1960 bytes\n",
            "Original size: 1146849 bytes \t Encoded size: 101396 bytes\n",
            "Original size: 649173 bytes \t Encoded size: 80456 bytes\n",
            "Original size: 251589 bytes \t Encoded size: 51048 bytes\n",
            "Original size: 562653 bytes \t Encoded size: 70200 bytes\n",
            "Original size: 90561 bytes \t Encoded size: 15044 bytes\n",
            "Original size: 285 bytes \t Encoded size: 160 bytes\n",
            "Original size: 1917 bytes \t Encoded size: 576 bytes\n",
            "Original size: 10497 bytes \t Encoded size: 2116 bytes\n",
            "Original size: 12213 bytes \t Encoded size: 2600 bytes\n",
            "Original size: 57 bytes \t Encoded size: 88 bytes\n",
            "Original size: 161073 bytes \t Encoded size: 20472 bytes\n",
            "Original size: 2733 bytes \t Encoded size: 908 bytes\n",
            "Number of labels: 17\n",
            "Original size: 104229 bytes \t Encoded size: 19612 bytes\n",
            "Original size: 183093 bytes \t Encoded size: 34284 bytes\n",
            "Original size: 83421 bytes \t Encoded size: 15636 bytes\n",
            "Original size: 4425 bytes \t Encoded size: 1180 bytes\n",
            "Original size: 1183689 bytes \t Encoded size: 121508 bytes\n",
            "Original size: 675249 bytes \t Encoded size: 87176 bytes\n",
            "Original size: 124749 bytes \t Encoded size: 27000 bytes\n",
            "Original size: 562653 bytes \t Encoded size: 70196 bytes\n",
            "Original size: 75213 bytes \t Encoded size: 14444 bytes\n",
            "Original size: 105 bytes \t Encoded size: 104 bytes\n",
            "Original size: 213 bytes \t Encoded size: 140 bytes\n",
            "Original size: 10761 bytes \t Encoded size: 2516 bytes\n",
            "Original size: 1257 bytes \t Encoded size: 408 bytes\n",
            "Original size: 30885 bytes \t Encoded size: 5372 bytes\n",
            "Original size: 5949 bytes \t Encoded size: 1332 bytes\n",
            "Original size: 65253 bytes \t Encoded size: 11052 bytes\n",
            "Original size: 4389 bytes \t Encoded size: 1012 bytes\n",
            "Number of labels: 17\n",
            "Original size: 165717 bytes \t Encoded size: 30924 bytes\n",
            "Original size: 365313 bytes \t Encoded size: 66524 bytes\n",
            "Original size: 45477 bytes \t Encoded size: 7232 bytes\n",
            "Original size: 13905 bytes \t Encoded size: 2472 bytes\n",
            "Original size: 1200561 bytes \t Encoded size: 102020 bytes\n",
            "Original size: 646005 bytes \t Encoded size: 78344 bytes\n",
            "Original size: 276741 bytes \t Encoded size: 53880 bytes\n",
            "Original size: 562653 bytes \t Encoded size: 70212 bytes\n",
            "Original size: 35025 bytes \t Encoded size: 7196 bytes\n",
            "Original size: 213 bytes \t Encoded size: 136 bytes\n",
            "Original size: 1761 bytes \t Encoded size: 496 bytes\n",
            "Original size: 20949 bytes \t Encoded size: 3288 bytes\n",
            "Original size: 56289 bytes \t Encoded size: 7296 bytes\n",
            "Original size: 1725 bytes \t Encoded size: 456 bytes\n",
            "Original size: 54093 bytes \t Encoded size: 9064 bytes\n",
            "Original size: 2853 bytes \t Encoded size: 1064 bytes\n",
            "Original size: 513 bytes \t Encoded size: 256 bytes\n",
            "Number of labels: 19\n",
            "Original size: 6957 bytes \t Encoded size: 1992 bytes\n",
            "Original size: 426261 bytes \t Encoded size: 66268 bytes\n",
            "Original size: 62961 bytes \t Encoded size: 9652 bytes\n",
            "Original size: 28425 bytes \t Encoded size: 3456 bytes\n",
            "Original size: 1161633 bytes \t Encoded size: 98716 bytes\n",
            "Original size: 616509 bytes \t Encoded size: 72668 bytes\n",
            "Original size: 201957 bytes \t Encoded size: 41128 bytes\n",
            "Original size: 562641 bytes \t Encoded size: 70208 bytes\n",
            "Original size: 1833 bytes \t Encoded size: 684 bytes\n",
            "Original size: 3381 bytes \t Encoded size: 604 bytes\n",
            "Original size: 1953 bytes \t Encoded size: 552 bytes\n",
            "Original size: 1953 bytes \t Encoded size: 520 bytes\n",
            "Original size: 9861 bytes \t Encoded size: 2008 bytes\n",
            "Original size: 81 bytes \t Encoded size: 92 bytes\n",
            "Original size: 61689 bytes \t Encoded size: 7776 bytes\n",
            "Original size: 28329 bytes \t Encoded size: 3756 bytes\n",
            "Original size: 78441 bytes \t Encoded size: 11816 bytes\n",
            "Original size: 1845 bytes \t Encoded size: 644 bytes\n",
            "Original size: 18933 bytes \t Encoded size: 2852 bytes\n",
            "Number of labels: 20\n",
            "Original size: 135381 bytes \t Encoded size: 28988 bytes\n",
            "Original size: 5493 bytes \t Encoded size: 1784 bytes\n",
            "Original size: 93 bytes \t Encoded size: 100 bytes\n",
            "Original size: 4437 bytes \t Encoded size: 1088 bytes\n",
            "Original size: 1011405 bytes \t Encoded size: 83512 bytes\n",
            "Original size: 341877 bytes \t Encoded size: 36368 bytes\n",
            "Original size: 214653 bytes \t Encoded size: 42504 bytes\n",
            "Original size: 562653 bytes \t Encoded size: 70212 bytes\n",
            "Original size: 2109 bytes \t Encoded size: 684 bytes\n",
            "Original size: 140301 bytes \t Encoded size: 25908 bytes\n",
            "Original size: 298293 bytes \t Encoded size: 35820 bytes\n",
            "Original size: 117 bytes \t Encoded size: 108 bytes\n",
            "Original size: 549 bytes \t Encoded size: 268 bytes\n",
            "Original size: 1029 bytes \t Encoded size: 424 bytes\n",
            "Original size: 2913 bytes \t Encoded size: 832 bytes\n",
            "Original size: 801 bytes \t Encoded size: 296 bytes\n",
            "Original size: 256281 bytes \t Encoded size: 26024 bytes\n",
            "Original size: 399069 bytes \t Encoded size: 86600 bytes\n",
            "Original size: 45 bytes \t Encoded size: 80 bytes\n",
            "Original size: 177 bytes \t Encoded size: 132 bytes\n",
            "Number of labels: 16\n",
            "Original size: 5361 bytes \t Encoded size: 1448 bytes\n",
            "Original size: 436425 bytes \t Encoded size: 67396 bytes\n",
            "Original size: 64929 bytes \t Encoded size: 10156 bytes\n",
            "Original size: 19965 bytes \t Encoded size: 2392 bytes\n",
            "Original size: 1190589 bytes \t Encoded size: 98916 bytes\n",
            "Original size: 603261 bytes \t Encoded size: 69504 bytes\n",
            "Original size: 243045 bytes \t Encoded size: 41932 bytes\n",
            "Original size: 562641 bytes \t Encoded size: 70192 bytes\n",
            "Original size: 9009 bytes \t Encoded size: 1668 bytes\n",
            "Original size: 69 bytes \t Encoded size: 92 bytes\n",
            "Original size: 81 bytes \t Encoded size: 96 bytes\n",
            "Original size: 81 bytes \t Encoded size: 96 bytes\n",
            "Original size: 18945 bytes \t Encoded size: 2968 bytes\n",
            "Original size: 4077 bytes \t Encoded size: 940 bytes\n",
            "Original size: 114549 bytes \t Encoded size: 14080 bytes\n",
            "Original size: 47841 bytes \t Encoded size: 7632 bytes\n",
            "Number of labels: 19\n",
            "Original size: 112449 bytes \t Encoded size: 22040 bytes\n",
            "Original size: 111309 bytes \t Encoded size: 25160 bytes\n",
            "Original size: 56529 bytes \t Encoded size: 10984 bytes\n",
            "Original size: 17937 bytes \t Encoded size: 3236 bytes\n",
            "Original size: 1350141 bytes \t Encoded size: 132396 bytes\n",
            "Original size: 359181 bytes \t Encoded size: 48976 bytes\n",
            "Original size: 263613 bytes \t Encoded size: 49916 bytes\n",
            "Original size: 562653 bytes \t Encoded size: 70184 bytes\n",
            "Original size: 16017 bytes \t Encoded size: 3656 bytes\n",
            "Original size: 165 bytes \t Encoded size: 124 bytes\n",
            "Original size: 209205 bytes \t Encoded size: 26128 bytes\n",
            "Original size: 1809 bytes \t Encoded size: 560 bytes\n",
            "Original size: 4653 bytes \t Encoded size: 1184 bytes\n",
            "Original size: 273 bytes \t Encoded size: 152 bytes\n",
            "Original size: 67017 bytes \t Encoded size: 10728 bytes\n",
            "Original size: 2433 bytes \t Encoded size: 644 bytes\n",
            "Original size: 101781 bytes \t Encoded size: 14192 bytes\n",
            "Original size: 2361 bytes \t Encoded size: 648 bytes\n",
            "Original size: 1857 bytes \t Encoded size: 584 bytes\n",
            "Number of labels: 20\n",
            "Original size: 417 bytes \t Encoded size: 220 bytes\n",
            "Original size: 576501 bytes \t Encoded size: 93388 bytes\n",
            "Original size: 8565 bytes \t Encoded size: 1880 bytes\n",
            "Original size: 9477 bytes \t Encoded size: 1784 bytes\n",
            "Original size: 1143261 bytes \t Encoded size: 112992 bytes\n",
            "Original size: 695937 bytes \t Encoded size: 90240 bytes\n",
            "Original size: 5781 bytes \t Encoded size: 1652 bytes\n",
            "Original size: 562653 bytes \t Encoded size: 70192 bytes\n",
            "Original size: 103365 bytes \t Encoded size: 19864 bytes\n",
            "Original size: 153 bytes \t Encoded size: 116 bytes\n",
            "Original size: 849 bytes \t Encoded size: 336 bytes\n",
            "Original size: 333 bytes \t Encoded size: 160 bytes\n",
            "Original size: 34593 bytes \t Encoded size: 6944 bytes\n",
            "Original size: 18189 bytes \t Encoded size: 3428 bytes\n",
            "Original size: 15789 bytes \t Encoded size: 2972 bytes\n",
            "Original size: 6045 bytes \t Encoded size: 1036 bytes\n",
            "Original size: 211101 bytes \t Encoded size: 23612 bytes\n",
            "Original size: 7329 bytes \t Encoded size: 2444 bytes\n",
            "Original size: 45 bytes \t Encoded size: 80 bytes\n",
            "Original size: 3741 bytes \t Encoded size: 1020 bytes\n",
            "Number of labels: 15\n",
            "Original size: 551265 bytes \t Encoded size: 88132 bytes\n",
            "Original size: 3081 bytes \t Encoded size: 1072 bytes\n",
            "Original size: 340257 bytes \t Encoded size: 54140 bytes\n",
            "Original size: 40005 bytes \t Encoded size: 4320 bytes\n",
            "Original size: 1175877 bytes \t Encoded size: 118244 bytes\n",
            "Original size: 558501 bytes \t Encoded size: 71516 bytes\n",
            "Original size: 31437 bytes \t Encoded size: 8112 bytes\n",
            "Original size: 562653 bytes \t Encoded size: 70172 bytes\n",
            "Original size: 633 bytes \t Encoded size: 284 bytes\n",
            "Original size: 141 bytes \t Encoded size: 112 bytes\n",
            "Original size: 57009 bytes \t Encoded size: 10820 bytes\n",
            "Original size: 405 bytes \t Encoded size: 196 bytes\n",
            "Original size: 141 bytes \t Encoded size: 116 bytes\n",
            "Original size: 23337 bytes \t Encoded size: 4072 bytes\n",
            "Original size: 121509 bytes \t Encoded size: 16832 bytes\n",
            "Number of labels: 14\n",
            "Original size: 877209 bytes \t Encoded size: 141368 bytes\n",
            "Original size: 22485 bytes \t Encoded size: 6344 bytes\n",
            "Original size: 11901 bytes \t Encoded size: 2104 bytes\n",
            "Original size: 1185657 bytes \t Encoded size: 108644 bytes\n",
            "Original size: 642621 bytes \t Encoded size: 78844 bytes\n",
            "Original size: 124677 bytes \t Encoded size: 25188 bytes\n",
            "Original size: 562653 bytes \t Encoded size: 70176 bytes\n",
            "Original size: 9789 bytes \t Encoded size: 2252 bytes\n",
            "Original size: 2805 bytes \t Encoded size: 940 bytes\n",
            "Original size: 19857 bytes \t Encoded size: 3716 bytes\n",
            "Original size: 6873 bytes \t Encoded size: 1324 bytes\n",
            "Original size: 14421 bytes \t Encoded size: 3040 bytes\n",
            "Original size: 669 bytes \t Encoded size: 312 bytes\n",
            "Original size: 3225 bytes \t Encoded size: 744 bytes\n",
            "Number of labels: 19\n",
            "Original size: 462477 bytes \t Encoded size: 81364 bytes\n",
            "Original size: 39753 bytes \t Encoded size: 9492 bytes\n",
            "Original size: 92109 bytes \t Encoded size: 14788 bytes\n",
            "Original size: 17265 bytes \t Encoded size: 3328 bytes\n",
            "Original size: 1512477 bytes \t Encoded size: 150780 bytes\n",
            "Original size: 343341 bytes \t Encoded size: 49044 bytes\n",
            "Original size: 194529 bytes \t Encoded size: 36432 bytes\n",
            "Original size: 562653 bytes \t Encoded size: 70188 bytes\n",
            "Original size: 14421 bytes \t Encoded size: 3412 bytes\n",
            "Original size: 69 bytes \t Encoded size: 92 bytes\n",
            "Original size: 105 bytes \t Encoded size: 108 bytes\n",
            "Original size: 129 bytes \t Encoded size: 112 bytes\n",
            "Original size: 2733 bytes \t Encoded size: 636 bytes\n",
            "Original size: 3285 bytes \t Encoded size: 888 bytes\n",
            "Original size: 4365 bytes \t Encoded size: 1164 bytes\n",
            "Original size: 48993 bytes \t Encoded size: 8704 bytes\n",
            "Original size: 45 bytes \t Encoded size: 80 bytes\n",
            "Original size: 108969 bytes \t Encoded size: 16452 bytes\n",
            "Original size: 177 bytes \t Encoded size: 136 bytes\n",
            "Number of labels: 20\n",
            "Original size: 120669 bytes \t Encoded size: 25488 bytes\n",
            "Original size: 155277 bytes \t Encoded size: 30204 bytes\n",
            "Original size: 51909 bytes \t Encoded size: 9112 bytes\n",
            "Original size: 15873 bytes \t Encoded size: 3076 bytes\n",
            "Original size: 1187961 bytes \t Encoded size: 102380 bytes\n",
            "Original size: 661989 bytes \t Encoded size: 77804 bytes\n",
            "Original size: 210921 bytes \t Encoded size: 47168 bytes\n",
            "Original size: 562653 bytes \t Encoded size: 70208 bytes\n",
            "Original size: 115449 bytes \t Encoded size: 19188 bytes\n",
            "Original size: 37101 bytes \t Encoded size: 8388 bytes\n",
            "Original size: 309 bytes \t Encoded size: 176 bytes\n",
            "Original size: 47397 bytes \t Encoded size: 6816 bytes\n",
            "Original size: 3873 bytes \t Encoded size: 776 bytes\n",
            "Original size: 14157 bytes \t Encoded size: 3556 bytes\n",
            "Original size: 20109 bytes \t Encoded size: 4096 bytes\n",
            "Original size: 1341 bytes \t Encoded size: 420 bytes\n",
            "Original size: 10377 bytes \t Encoded size: 2496 bytes\n",
            "Original size: 501 bytes \t Encoded size: 244 bytes\n",
            "Original size: 9909 bytes \t Encoded size: 2904 bytes\n",
            "Original size: 585 bytes \t Encoded size: 256 bytes\n",
            "Number of labels: 18\n",
            "Original size: 188337 bytes \t Encoded size: 31496 bytes\n",
            "Original size: 149901 bytes \t Encoded size: 34984 bytes\n",
            "Original size: 56565 bytes \t Encoded size: 11092 bytes\n",
            "Original size: 22701 bytes \t Encoded size: 3560 bytes\n",
            "Original size: 1322517 bytes \t Encoded size: 125088 bytes\n",
            "Original size: 544113 bytes \t Encoded size: 68808 bytes\n",
            "Original size: 262305 bytes \t Encoded size: 46604 bytes\n",
            "Original size: 562653 bytes \t Encoded size: 70200 bytes\n",
            "Original size: 21873 bytes \t Encoded size: 5068 bytes\n",
            "Original size: 225 bytes \t Encoded size: 140 bytes\n",
            "Original size: 30069 bytes \t Encoded size: 5508 bytes\n",
            "Original size: 2145 bytes \t Encoded size: 584 bytes\n",
            "Original size: 4785 bytes \t Encoded size: 1224 bytes\n",
            "Original size: 429 bytes \t Encoded size: 200 bytes\n",
            "Original size: 39753 bytes \t Encoded size: 7124 bytes\n",
            "Original size: 19221 bytes \t Encoded size: 3048 bytes\n",
            "Original size: 17961 bytes \t Encoded size: 3932 bytes\n",
            "Original size: 11361 bytes \t Encoded size: 2256 bytes\n",
            "Number of labels: 18\n",
            "Original size: 4101 bytes \t Encoded size: 1144 bytes\n",
            "Original size: 193161 bytes \t Encoded size: 38856 bytes\n",
            "Original size: 44577 bytes \t Encoded size: 8352 bytes\n",
            "Original size: 12801 bytes \t Encoded size: 2440 bytes\n",
            "Original size: 1196841 bytes \t Encoded size: 102624 bytes\n",
            "Original size: 620433 bytes \t Encoded size: 72440 bytes\n",
            "Original size: 151245 bytes \t Encoded size: 30364 bytes\n",
            "Original size: 562653 bytes \t Encoded size: 70184 bytes\n",
            "Original size: 267885 bytes \t Encoded size: 38644 bytes\n",
            "Original size: 3213 bytes \t Encoded size: 836 bytes\n",
            "Original size: 213 bytes \t Encoded size: 144 bytes\n",
            "Original size: 333 bytes \t Encoded size: 180 bytes\n",
            "Original size: 118485 bytes \t Encoded size: 16688 bytes\n",
            "Original size: 3825 bytes \t Encoded size: 1356 bytes\n",
            "Original size: 77877 bytes \t Encoded size: 10472 bytes\n",
            "Original size: 45 bytes \t Encoded size: 80 bytes\n",
            "Original size: 6381 bytes \t Encoded size: 1432 bytes\n",
            "Original size: 25449 bytes \t Encoded size: 5284 bytes\n"
          ]
        }
      ],
      "source": [
        "# go through all the files and save the results\n",
        "draco_labeled_res = dict()\n",
        "for file in all_pc:\n",
        "    draco_labeled_res[file] = draco_labeled(file)"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 17,
      "metadata": {},
      "outputs": [],
      "source": [
        "# read files \n",
        "draco_labeled_res = dict()\n",
        "for file in all_pc:\n",
        "    img_name = file.split('/')[-1].split('.')[0]\n",
        "    res_dir = 'results'\n",
        "    draco_labeled_res[file] = pd.read_csv(os.path.join(res_dir, img_name+'draco_labeled.csv'))\n"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 18,
      "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>label</th>\n",
              "      <th>original_size</th>\n",
              "      <th>encoded_size</th>\n",
              "      <th>encoding_decoding_time</th>\n",
              "      <th>lost_points</th>\n",
              "      <th>mpde</th>\n",
              "      <th>compression_ratio</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>0.0</td>\n",
              "      <td>274077.0</td>\n",
              "      <td>1144.0</td>\n",
              "      <td>0.003294</td>\n",
              "      <td>273738.0</td>\n",
              "      <td>0.003797</td>\n",
              "      <td>0.278956</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1</th>\n",
              "      <td>1.0</td>\n",
              "      <td>274077.0</td>\n",
              "      <td>38856.0</td>\n",
              "      <td>0.049923</td>\n",
              "      <td>257984.0</td>\n",
              "      <td>0.003774</td>\n",
              "      <td>0.201159</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2</th>\n",
              "      <td>2.0</td>\n",
              "      <td>274077.0</td>\n",
              "      <td>8352.0</td>\n",
              "      <td>0.010851</td>\n",
              "      <td>270365.0</td>\n",
              "      <td>0.005662</td>\n",
              "      <td>0.187361</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3</th>\n",
              "      <td>5.0</td>\n",
              "      <td>274077.0</td>\n",
              "      <td>2440.0</td>\n",
              "      <td>0.004791</td>\n",
              "      <td>273013.0</td>\n",
              "      <td>0.005171</td>\n",
              "      <td>0.190610</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4</th>\n",
              "      <td>7.0</td>\n",
              "      <td>274077.0</td>\n",
              "      <td>102624.0</td>\n",
              "      <td>0.312609</td>\n",
              "      <td>174783.0</td>\n",
              "      <td>0.003533</td>\n",
              "      <td>0.085746</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>5</th>\n",
              "      <td>8.0</td>\n",
              "      <td>274077.0</td>\n",
              "      <td>72440.0</td>\n",
              "      <td>0.162096</td>\n",
              "      <td>222403.0</td>\n",
              "      <td>0.004278</td>\n",
              "      <td>0.116757</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>6</th>\n",
              "      <td>9.0</td>\n",
              "      <td>274077.0</td>\n",
              "      <td>30364.0</td>\n",
              "      <td>0.051827</td>\n",
              "      <td>261476.0</td>\n",
              "      <td>0.005247</td>\n",
              "      <td>0.200760</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>7</th>\n",
              "      <td>10.0</td>\n",
              "      <td>274077.0</td>\n",
              "      <td>70184.0</td>\n",
              "      <td>0.131092</td>\n",
              "      <td>227192.0</td>\n",
              "      <td>0.000159</td>\n",
              "      <td>0.124738</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>8</th>\n",
              "      <td>11.0</td>\n",
              "      <td>274077.0</td>\n",
              "      <td>38644.0</td>\n",
              "      <td>0.058764</td>\n",
              "      <td>251756.0</td>\n",
              "      <td>0.002873</td>\n",
              "      <td>0.144256</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>9</th>\n",
              "      <td>12.0</td>\n",
              "      <td>274077.0</td>\n",
              "      <td>836.0</td>\n",
              "      <td>0.003122</td>\n",
              "      <td>273812.0</td>\n",
              "      <td>0.000685</td>\n",
              "      <td>0.260193</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>10</th>\n",
              "      <td>17.0</td>\n",
              "      <td>274077.0</td>\n",
              "      <td>144.0</td>\n",
              "      <td>0.001490</td>\n",
              "      <td>274062.0</td>\n",
              "      <td>0.000172</td>\n",
              "      <td>0.676056</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>11</th>\n",
              "      <td>18.0</td>\n",
              "      <td>274077.0</td>\n",
              "      <td>180.0</td>\n",
              "      <td>0.001290</td>\n",
              "      <td>274052.0</td>\n",
              "      <td>0.000693</td>\n",
              "      <td>0.540541</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>12</th>\n",
              "      <td>19.0</td>\n",
              "      <td>274077.0</td>\n",
              "      <td>16688.0</td>\n",
              "      <td>0.027093</td>\n",
              "      <td>264206.0</td>\n",
              "      <td>0.003831</td>\n",
              "      <td>0.140845</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>13</th>\n",
              "      <td>20.0</td>\n",
              "      <td>274077.0</td>\n",
              "      <td>1356.0</td>\n",
              "      <td>0.002389</td>\n",
              "      <td>273761.0</td>\n",
              "      <td>0.000099</td>\n",
              "      <td>0.354510</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>14</th>\n",
              "      <td>40.0</td>\n",
              "      <td>274077.0</td>\n",
              "      <td>10472.0</td>\n",
              "      <td>0.018137</td>\n",
              "      <td>267590.0</td>\n",
              "      <td>0.005384</td>\n",
              "      <td>0.134468</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>15</th>\n",
              "      <td>41.0</td>\n",
              "      <td>274077.0</td>\n",
              "      <td>80.0</td>\n",
              "      <td>0.002053</td>\n",
              "      <td>274076.0</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>1.777778</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>16</th>\n",
              "      <td>100.0</td>\n",
              "      <td>274077.0</td>\n",
              "      <td>1432.0</td>\n",
              "      <td>0.002978</td>\n",
              "      <td>273548.0</td>\n",
              "      <td>0.004162</td>\n",
              "      <td>0.224416</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>17</th>\n",
              "      <td>105.0</td>\n",
              "      <td>274077.0</td>\n",
              "      <td>5284.0</td>\n",
              "      <td>0.009932</td>\n",
              "      <td>271959.0</td>\n",
              "      <td>0.000775</td>\n",
              "      <td>0.207631</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>"
            ],
            "text/plain": [
              "    label  original_size  encoded_size  encoding_decoding_time  lost_points  \\\n",
              "0     0.0       274077.0        1144.0                0.003294     273738.0   \n",
              "1     1.0       274077.0       38856.0                0.049923     257984.0   \n",
              "2     2.0       274077.0        8352.0                0.010851     270365.0   \n",
              "3     5.0       274077.0        2440.0                0.004791     273013.0   \n",
              "4     7.0       274077.0      102624.0                0.312609     174783.0   \n",
              "5     8.0       274077.0       72440.0                0.162096     222403.0   \n",
              "6     9.0       274077.0       30364.0                0.051827     261476.0   \n",
              "7    10.0       274077.0       70184.0                0.131092     227192.0   \n",
              "8    11.0       274077.0       38644.0                0.058764     251756.0   \n",
              "9    12.0       274077.0         836.0                0.003122     273812.0   \n",
              "10   17.0       274077.0         144.0                0.001490     274062.0   \n",
              "11   18.0       274077.0         180.0                0.001290     274052.0   \n",
              "12   19.0       274077.0       16688.0                0.027093     264206.0   \n",
              "13   20.0       274077.0        1356.0                0.002389     273761.0   \n",
              "14   40.0       274077.0       10472.0                0.018137     267590.0   \n",
              "15   41.0       274077.0          80.0                0.002053     274076.0   \n",
              "16  100.0       274077.0        1432.0                0.002978     273548.0   \n",
              "17  105.0       274077.0        5284.0                0.009932     271959.0   \n",
              "\n",
              "        mpde  compression_ratio  \n",
              "0   0.003797           0.278956  \n",
              "1   0.003774           0.201159  \n",
              "2   0.005662           0.187361  \n",
              "3   0.005171           0.190610  \n",
              "4   0.003533           0.085746  \n",
              "5   0.004278           0.116757  \n",
              "6   0.005247           0.200760  \n",
              "7   0.000159           0.124738  \n",
              "8   0.002873           0.144256  \n",
              "9   0.000685           0.260193  \n",
              "10  0.000172           0.676056  \n",
              "11  0.000693           0.540541  \n",
              "12  0.003831           0.140845  \n",
              "13  0.000099           0.354510  \n",
              "14  0.005384           0.134468  \n",
              "15  0.000000           1.777778  \n",
              "16  0.004162           0.224416  \n",
              "17  0.000775           0.207631  "
            ]
          },
          "execution_count": 18,
          "metadata": {},
          "output_type": "execute_result"
        }
      ],
      "source": [
        "draco_labeled_res[file]"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 19,
      "metadata": {},
      "outputs": [],
      "source": [
        "# stats for each label\n",
        "stats_labels = pd.DataFrame(columns=['label', \n",
        "                                    'avg_compression_ratio', \n",
        "                                    'std_compression_ratio', \n",
        "                                    'avg_encoding_decoding_time', \n",
        "                                    'std_encoding_decoding_time', \n",
        "                                    'avg_lost_points/total_points', \n",
        "                                    'std_lost_points/total_points',\n",
        "                                    'avg_mpde',\n",
        "                                    ])\n",
        "# for each label plot the avg compression ratio and the avg encoding-decoding time\n",
        "\n",
        "all_labels = np.unique(np.concatenate([draco_labeled_res[file]['label'].values for file in all_pc]))\n",
        "for label in all_labels:\n",
        "    avg_compression_ratio = np.mean([draco_labeled_res[file][draco_labeled_res[file]['label'] == label]['compression_ratio'].iloc[0] for file in all_pc if label in draco_labeled_res[file]['label'].values])\n",
        "    std_compression_ratio = np.std([draco_labeled_res[file][draco_labeled_res[file]['label'] == label]['compression_ratio'].iloc[0] for file in all_pc if label in draco_labeled_res[file]['label'].values])\n",
        "    avg_encoding_decoding_time = np.mean([draco_labeled_res[file][draco_labeled_res[file]['label'] == label]['encoding_decoding_time'].iloc[0] for file in all_pc if label in draco_labeled_res[file]['label'].values])\n",
        "    std_encoding_decoding_time = np.std([draco_labeled_res[file][draco_labeled_res[file]['label'] == label]['encoding_decoding_time'].iloc[0] for file in all_pc if label in draco_labeled_res[file]['label'].values])\n",
        "    avg_lost_points = np.mean([draco_labeled_res[file][draco_labeled_res[file]['label'] == label]['lost_points']/draco_labeled_res[file][draco_labeled_res[file]['label'] == label]['original_size'].iloc[0]  for file in all_pc if label in draco_labeled_res[file]['label'].values])\n",
        "    std_lost_points = np.std([draco_labeled_res[file][draco_labeled_res[file]['label'] == label]['lost_points']/draco_labeled_res[file][draco_labeled_res[file]['label'] == label]['original_size'].iloc[0]  for file in all_pc if label in draco_labeled_res[file]['label'].values])\n",
        "    avg_mpde = np.mean([draco_labeled_res[file][draco_labeled_res[file]['label'] == label]['mpde'].iloc[0] for file in all_pc if label in draco_labeled_res[file]['label'].values])\n",
        "\n",
        "    stats_labels = stats_labels.append({'label': label,\n",
        "                                        'avg_compression_ratio': avg_compression_ratio, \n",
        "                                        'std_compression_ratio': std_compression_ratio,\n",
        "                                        'avg_encoding_decoding_time': avg_encoding_decoding_time,\n",
        "                                        'std_encoding_decoding_time': std_encoding_decoding_time,\n",
        "                                        'avg_lost_points/total_points': avg_lost_points,\n",
        "                                        'std_lost_points/total_points': std_lost_points,\n",
        "                                        'avg_mpde': avg_mpde\n",
        "                                        },\n",
        "                                        ignore_index=True)"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 20,
      "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>label</th>\n",
              "      <th>avg_compression_ratio</th>\n",
              "      <th>std_compression_ratio</th>\n",
              "      <th>avg_encoding_decoding_time</th>\n",
              "      <th>std_encoding_decoding_time</th>\n",
              "      <th>avg_lost_points/total_points</th>\n",
              "      <th>std_lost_points/total_points</th>\n",
              "      <th>avg_mpde</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>0.0</td>\n",
              "      <td>0.213979</td>\n",
              "      <td>0.065690</td>\n",
              "      <td>0.046912</td>\n",
              "      <td>0.051920</td>\n",
              "      <td>0.949713</td>\n",
              "      <td>0.057275</td>\n",
              "      <td>0.004285</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1</th>\n",
              "      <td>1.0</td>\n",
              "      <td>0.207564</td>\n",
              "      <td>0.045764</td>\n",
              "      <td>0.055434</td>\n",
              "      <td>0.041921</td>\n",
              "      <td>0.939731</td>\n",
              "      <td>0.043845</td>\n",
              "      <td>0.004182</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2</th>\n",
              "      <td>2.0</td>\n",
              "      <td>0.215043</td>\n",
              "      <td>0.100345</td>\n",
              "      <td>0.013153</td>\n",
              "      <td>0.013533</td>\n",
              "      <td>0.985899</td>\n",
              "      <td>0.017724</td>\n",
              "      <td>0.004154</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3</th>\n",
              "      <td>3.0</td>\n",
              "      <td>0.724464</td>\n",
              "      <td>0.390792</td>\n",
              "      <td>0.001846</td>\n",
              "      <td>0.000510</td>\n",
              "      <td>0.999755</td>\n",
              "      <td>0.000449</td>\n",
              "      <td>0.000327</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4</th>\n",
              "      <td>5.0</td>\n",
              "      <td>0.178832</td>\n",
              "      <td>0.037474</td>\n",
              "      <td>0.005670</td>\n",
              "      <td>0.002732</td>\n",
              "      <td>0.994832</td>\n",
              "      <td>0.002975</td>\n",
              "      <td>0.004658</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>5</th>\n",
              "      <td>7.0</td>\n",
              "      <td>0.090361</td>\n",
              "      <td>0.006316</td>\n",
              "      <td>0.329025</td>\n",
              "      <td>0.057278</td>\n",
              "      <td>0.634253</td>\n",
              "      <td>0.044794</td>\n",
              "      <td>0.003427</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>6</th>\n",
              "      <td>8.0</td>\n",
              "      <td>0.124917</td>\n",
              "      <td>0.010122</td>\n",
              "      <td>0.155289</td>\n",
              "      <td>0.038902</td>\n",
              "      <td>0.823944</td>\n",
              "      <td>0.038224</td>\n",
              "      <td>0.003688</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>7</th>\n",
              "      <td>9.0</td>\n",
              "      <td>0.217767</td>\n",
              "      <td>0.117306</td>\n",
              "      <td>0.038156</td>\n",
              "      <td>0.019669</td>\n",
              "      <td>0.954100</td>\n",
              "      <td>0.024590</td>\n",
              "      <td>0.004989</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>8</th>\n",
              "      <td>10.0</td>\n",
              "      <td>0.124748</td>\n",
              "      <td>0.000021</td>\n",
              "      <td>0.141588</td>\n",
              "      <td>0.016786</td>\n",
              "      <td>0.828065</td>\n",
              "      <td>0.006561</td>\n",
              "      <td>0.000159</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>9</th>\n",
              "      <td>11.0</td>\n",
              "      <td>0.227799</td>\n",
              "      <td>0.082529</td>\n",
              "      <td>0.015390</td>\n",
              "      <td>0.015277</td>\n",
              "      <td>0.984710</td>\n",
              "      <td>0.018950</td>\n",
              "      <td>0.003460</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>10</th>\n",
              "      <td>12.0</td>\n",
              "      <td>0.735115</td>\n",
              "      <td>0.503065</td>\n",
              "      <td>0.002184</td>\n",
              "      <td>0.001161</td>\n",
              "      <td>0.999641</td>\n",
              "      <td>0.000543</td>\n",
              "      <td>0.000728</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>11</th>\n",
              "      <td>15.0</td>\n",
              "      <td>0.415259</td>\n",
              "      <td>0.458475</td>\n",
              "      <td>0.014904</td>\n",
              "      <td>0.021064</td>\n",
              "      <td>0.982442</td>\n",
              "      <td>0.027719</td>\n",
              "      <td>0.002237</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>12</th>\n",
              "      <td>17.0</td>\n",
              "      <td>0.382645</td>\n",
              "      <td>0.295025</td>\n",
              "      <td>0.010771</td>\n",
              "      <td>0.018186</td>\n",
              "      <td>0.987601</td>\n",
              "      <td>0.023943</td>\n",
              "      <td>0.001388</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>13</th>\n",
              "      <td>18.0</td>\n",
              "      <td>0.524343</td>\n",
              "      <td>0.458984</td>\n",
              "      <td>0.001968</td>\n",
              "      <td>0.000457</td>\n",
              "      <td>0.999546</td>\n",
              "      <td>0.000348</td>\n",
              "      <td>0.002141</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>14</th>\n",
              "      <td>19.0</td>\n",
              "      <td>0.255264</td>\n",
              "      <td>0.195003</td>\n",
              "      <td>0.007335</td>\n",
              "      <td>0.009060</td>\n",
              "      <td>0.992477</td>\n",
              "      <td>0.009527</td>\n",
              "      <td>0.003158</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>15</th>\n",
              "      <td>20.0</td>\n",
              "      <td>0.376955</td>\n",
              "      <td>0.267165</td>\n",
              "      <td>0.003016</td>\n",
              "      <td>0.002817</td>\n",
              "      <td>0.998152</td>\n",
              "      <td>0.003431</td>\n",
              "      <td>0.002558</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>16</th>\n",
              "      <td>21.0</td>\n",
              "      <td>0.373665</td>\n",
              "      <td>0.070440</td>\n",
              "      <td>0.002410</td>\n",
              "      <td>0.001299</td>\n",
              "      <td>0.999038</td>\n",
              "      <td>0.001303</td>\n",
              "      <td>0.001491</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>17</th>\n",
              "      <td>40.0</td>\n",
              "      <td>0.161903</td>\n",
              "      <td>0.036578</td>\n",
              "      <td>0.012460</td>\n",
              "      <td>0.006683</td>\n",
              "      <td>0.985153</td>\n",
              "      <td>0.008057</td>\n",
              "      <td>0.004768</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>18</th>\n",
              "      <td>41.0</td>\n",
              "      <td>0.412539</td>\n",
              "      <td>0.370558</td>\n",
              "      <td>0.002608</td>\n",
              "      <td>0.001437</td>\n",
              "      <td>0.998672</td>\n",
              "      <td>0.001930</td>\n",
              "      <td>0.002464</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>19</th>\n",
              "      <td>100.0</td>\n",
              "      <td>0.144662</td>\n",
              "      <td>0.036092</td>\n",
              "      <td>0.032281</td>\n",
              "      <td>0.026846</td>\n",
              "      <td>0.958732</td>\n",
              "      <td>0.032902</td>\n",
              "      <td>0.004330</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>20</th>\n",
              "      <td>101.0</td>\n",
              "      <td>0.322917</td>\n",
              "      <td>0.118224</td>\n",
              "      <td>0.013310</td>\n",
              "      <td>0.025806</td>\n",
              "      <td>0.985622</td>\n",
              "      <td>0.031362</td>\n",
              "      <td>0.001090</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>21</th>\n",
              "      <td>102.0</td>\n",
              "      <td>0.358440</td>\n",
              "      <td>0.169791</td>\n",
              "      <td>0.025390</td>\n",
              "      <td>0.053780</td>\n",
              "      <td>0.973558</td>\n",
              "      <td>0.054686</td>\n",
              "      <td>0.000665</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>22</th>\n",
              "      <td>103.0</td>\n",
              "      <td>0.371783</td>\n",
              "      <td>0.269854</td>\n",
              "      <td>0.005591</td>\n",
              "      <td>0.010099</td>\n",
              "      <td>0.995131</td>\n",
              "      <td>0.013423</td>\n",
              "      <td>0.001417</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>23</th>\n",
              "      <td>104.0</td>\n",
              "      <td>0.762646</td>\n",
              "      <td>0.513707</td>\n",
              "      <td>0.002033</td>\n",
              "      <td>0.000799</td>\n",
              "      <td>0.999617</td>\n",
              "      <td>0.000832</td>\n",
              "      <td>0.000828</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>24</th>\n",
              "      <td>105.0</td>\n",
              "      <td>0.430213</td>\n",
              "      <td>0.321163</td>\n",
              "      <td>0.002429</td>\n",
              "      <td>0.001296</td>\n",
              "      <td>0.998880</td>\n",
              "      <td>0.001576</td>\n",
              "      <td>0.002148</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>"
            ],
            "text/plain": [
              "    label  avg_compression_ratio  std_compression_ratio  \\\n",
              "0     0.0               0.213979               0.065690   \n",
              "1     1.0               0.207564               0.045764   \n",
              "2     2.0               0.215043               0.100345   \n",
              "3     3.0               0.724464               0.390792   \n",
              "4     5.0               0.178832               0.037474   \n",
              "5     7.0               0.090361               0.006316   \n",
              "6     8.0               0.124917               0.010122   \n",
              "7     9.0               0.217767               0.117306   \n",
              "8    10.0               0.124748               0.000021   \n",
              "9    11.0               0.227799               0.082529   \n",
              "10   12.0               0.735115               0.503065   \n",
              "11   15.0               0.415259               0.458475   \n",
              "12   17.0               0.382645               0.295025   \n",
              "13   18.0               0.524343               0.458984   \n",
              "14   19.0               0.255264               0.195003   \n",
              "15   20.0               0.376955               0.267165   \n",
              "16   21.0               0.373665               0.070440   \n",
              "17   40.0               0.161903               0.036578   \n",
              "18   41.0               0.412539               0.370558   \n",
              "19  100.0               0.144662               0.036092   \n",
              "20  101.0               0.322917               0.118224   \n",
              "21  102.0               0.358440               0.169791   \n",
              "22  103.0               0.371783               0.269854   \n",
              "23  104.0               0.762646               0.513707   \n",
              "24  105.0               0.430213               0.321163   \n",
              "\n",
              "    avg_encoding_decoding_time  std_encoding_decoding_time  \\\n",
              "0                     0.046912                    0.051920   \n",
              "1                     0.055434                    0.041921   \n",
              "2                     0.013153                    0.013533   \n",
              "3                     0.001846                    0.000510   \n",
              "4                     0.005670                    0.002732   \n",
              "5                     0.329025                    0.057278   \n",
              "6                     0.155289                    0.038902   \n",
              "7                     0.038156                    0.019669   \n",
              "8                     0.141588                    0.016786   \n",
              "9                     0.015390                    0.015277   \n",
              "10                    0.002184                    0.001161   \n",
              "11                    0.014904                    0.021064   \n",
              "12                    0.010771                    0.018186   \n",
              "13                    0.001968                    0.000457   \n",
              "14                    0.007335                    0.009060   \n",
              "15                    0.003016                    0.002817   \n",
              "16                    0.002410                    0.001299   \n",
              "17                    0.012460                    0.006683   \n",
              "18                    0.002608                    0.001437   \n",
              "19                    0.032281                    0.026846   \n",
              "20                    0.013310                    0.025806   \n",
              "21                    0.025390                    0.053780   \n",
              "22                    0.005591                    0.010099   \n",
              "23                    0.002033                    0.000799   \n",
              "24                    0.002429                    0.001296   \n",
              "\n",
              "    avg_lost_points/total_points  std_lost_points/total_points  avg_mpde  \n",
              "0                       0.949713                      0.057275  0.004285  \n",
              "1                       0.939731                      0.043845  0.004182  \n",
              "2                       0.985899                      0.017724  0.004154  \n",
              "3                       0.999755                      0.000449  0.000327  \n",
              "4                       0.994832                      0.002975  0.004658  \n",
              "5                       0.634253                      0.044794  0.003427  \n",
              "6                       0.823944                      0.038224  0.003688  \n",
              "7                       0.954100                      0.024590  0.004989  \n",
              "8                       0.828065                      0.006561  0.000159  \n",
              "9                       0.984710                      0.018950  0.003460  \n",
              "10                      0.999641                      0.000543  0.000728  \n",
              "11                      0.982442                      0.027719  0.002237  \n",
              "12                      0.987601                      0.023943  0.001388  \n",
              "13                      0.999546                      0.000348  0.002141  \n",
              "14                      0.992477                      0.009527  0.003158  \n",
              "15                      0.998152                      0.003431  0.002558  \n",
              "16                      0.999038                      0.001303  0.001491  \n",
              "17                      0.985153                      0.008057  0.004768  \n",
              "18                      0.998672                      0.001930  0.002464  \n",
              "19                      0.958732                      0.032902  0.004330  \n",
              "20                      0.985622                      0.031362  0.001090  \n",
              "21                      0.973558                      0.054686  0.000665  \n",
              "22                      0.995131                      0.013423  0.001417  \n",
              "23                      0.999617                      0.000832  0.000828  \n",
              "24                      0.998880                      0.001576  0.002148  "
            ]
          },
          "execution_count": 20,
          "metadata": {},
          "output_type": "execute_result"
        }
      ],
      "source": [
        "stats_labels"
      ]
    },
    {
      "attachments": {},
      "cell_type": "markdown",
      "metadata": {},
      "source": [
        "Compression ratio is computed as "
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 29,
      "metadata": {},
      "outputs": [
        {
          "data": {
            "image/png": "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",
            "text/plain": [
              "<Figure size 1008x576 with 2 Axes>"
            ]
          },
          "metadata": {},
          "output_type": "display_data"
        }
      ],
      "source": [
        "# plot the results\n",
        "import seaborn as sns\n",
        "fig, ax = plt.subplots(1, 2, figsize=(14, 8))\n",
        "\n",
        "sns.set_theme(style=\"whitegrid\", font_scale=1.5)\n",
        "# add error bars\n",
        "# ax[0].errorbar(stats_labels['avg_lost_points/total_points'], y=stats_labels['avg_compression_ratio'], yerr=stats_labels['std_compression_ratio'], fmt='o', color='black')\n",
        "sns.pointplot(x='avg_mpde', y='avg_compression_ratio', data=stats_labels,  \n",
        "                ax=ax[0], dodge=True, join=False, palette='Set2', hue='label')\n",
        "\n",
        "\n",
        "ax[0].set_xlabel('avg mpde')\n",
        "ax[0].set_ylabel('avg compression ratio')\n",
        "ax[0].set_title('Compression ratio')\n",
        "# set xticks with double precision\n",
        "from matplotlib.ticker import FormatStrFormatter \n",
        "ax[0].xaxis.set_major_formatter(FormatStrFormatter('%.2f'))\n",
        "\n",
        "# print xticks every 5 ticks\n",
        "for i, label in enumerate(ax[0].get_xticklabels()):\n",
        "    if i % 5 != 0:\n",
        "        label.set_visible(False)\n",
        "    \n",
        "\n",
        "#remove the legend\n",
        "ax[0].legend_.remove()\n",
        "\n",
        "sns.pointplot(x='avg_mpde', y='avg_encoding_decoding_time', data=stats_labels,  ax=ax[1], dodge=True, join=False, palette='Set2', hue='label')\n",
        "ax[1].set_xlabel('avg mpde')\n",
        "ax[1].set_ylabel('avg encoding-decoding time')\n",
        "ax[1].set_title('Computational time')\n",
        "# put legend outside the plot\n",
        "ax[1].legend(loc='center left', bbox_to_anchor=(1, 0.5), ncol=1)\n",
        "\n",
        "# set xticks with double precision\n",
        "from matplotlib.ticker import FormatStrFormatter\n",
        "ax[1].xaxis.set_major_formatter(FormatStrFormatter('%.2f'))\n",
        "\n",
        "# print xticks every 5 ticks\n",
        "for i, label in enumerate(ax[1].get_xticklabels()):\n",
        "    if i % 5 != 0:\n",
        "        label.set_visible(False)\n",
        "\n",
        "plt.show()"
      ]
    }
  ],
  "metadata": {
    "colab": {
      "provenance": []
    },
    "kernelspec": {
      "display_name": "3d",
      "language": "python",
      "name": "3d"
    },
    "language_info": {
      "codemirror_mode": {
        "name": "ipython",
        "version": 3
      },
      "file_extension": ".py",
      "mimetype": "text/x-python",
      "name": "python",
      "nbconvert_exporter": "python",
      "pygments_lexer": "ipython3",
      "version": "3.9.13"
    }
  },
  "nbformat": 4,
  "nbformat_minor": 0
}