{
  "id": 456224,
  "title": "kidney5==kidney6 ?",
  "url": "/competitions/blood-vessel-segmentation/discussion/456224",
  "author_name": "",
  "post_date": "2023-11-18T20:05:32.262372900Z",
  "votes": 1,
  "comment_count": 2,
  "views": 0,
  "content": "<p>Hi,</p>\n<p>Just looking at the data provided.</p>\n<p>Kidney 5 and 6 images look exactly the same. Is this meant to be like that?</p>",
  "messages": [
    {
      "id": "2530128",
      "postDate": "11/18/2023 20:05:32",
      "content": "<p>Hi,</p>\n<p>Just looking at the data provided.</p>\n<p>Kidney 5 and 6 images look exactly the same. Is this meant to be like that?</p>",
      "rawMarkdown": "Hi,\n\nJust looking at the data provided.\n\nKidney 5 and 6 images look exactly the same. Is this meant to be like that?",
      "votes": null
    },
    {
      "id": "2530134",
      "postDate": "11/18/2023 20:13:15",
      "content": "<p><a href=\"https://www.kaggle.com/perdigao1/kidney5-kidney6\" target=\"_blank\">https://www.kaggle.com/perdigao1/kidney5-kidney6</a></p>",
      "rawMarkdown": "https://www.kaggle.com/perdigao1/kidney5-kidney6",
      "votes": null
    },
    {
      "id": "2530549",
      "postDate": "11/19/2023 08:45:11",
      "content": "<p>The data in the test folder are just dummy values. You should only care about how the test data is structured.<br>\nWhen reading the test data during inference, make sure that it is a generic way to extract all the TIFF files with the given test folder structure. Example:</p>\n<pre><code> numpy  np \n pandas  pd \n glob  glob\n os.path  basename, join\n\nDATASET_FOLDER = \ntest_image_filenames = glob(join(DATASET_FOLDER, , , , ))\n\nids = []\nrles = []\n filename  test_image_filenames:\n    label = filename.split()[]\n    slice_number = basename(filename).split()[]\n    ids.append()\n    predicted_rle = \n    rles.append(predicted_rle)\n\ndf = pd.DataFrame({: ids, : rles}).sort_values(by=[])\ndf.to_csv(, index=)\n</code></pre>",
      "rawMarkdown": "The data in the test folder are just dummy values. You should only care about how the test data is structured.\nWhen reading the test data during inference, make sure that it is a generic way to extract all the TIFF files with the given test folder structure. Example:\n```python\nimport numpy as np # linear algebra\nimport pandas as pd # data processing, CSV file I/O (e.g. pd.read_csv)\nfrom glob import glob\nfrom os.path import basename, join\n\nDATASET_FOLDER = \"/kaggle/input/blood-vessel-segmentation\"\ntest_image_filenames = glob(join(DATASET_FOLDER, \"test\", \"*\", \"*\", \"*.tif\"))\n\nids = []\nrles = []\nfor filename in test_image_filenames:\n    label = filename.split('/')[5]\n    slice_number = basename(filename).split('.')[0]\n    ids.append(f\"{label}_{slice_number}\")\n    predicted_rle = # Your inference code\n    rles.append(predicted_rle)\n\ndf = pd.DataFrame({'id': ids, 'rle': rles}).sort_values(by=['id'])\ndf.to_csv('submission.csv', index=False)\n```",
      "votes": null
    }
  ],
  "comments": [
    {
      "id": 2530134,
      "author_name": "perdigao1",
      "author_url": "",
      "post_date": "11/18/2023 20:13:15",
      "content": "<p><a href=\"https://www.kaggle.com/perdigao1/kidney5-kidney6\" target=\"_blank\">https://www.kaggle.com/perdigao1/kidney5-kidney6</a></p>",
      "votes": null,
      "replies": []
    },
    {
      "id": 2530549,
      "author_name": "coderrkj",
      "author_url": "",
      "post_date": "11/19/2023 08:45:11",
      "content": "<p>The data in the test folder are just dummy values. You should only care about how the test data is structured.<br>\nWhen reading the test data during inference, make sure that it is a generic way to extract all the TIFF files with the given test folder structure. Example:</p>\n<pre><code> numpy  np \n pandas  pd \n glob  glob\n os.path  basename, join\n\nDATASET_FOLDER = \ntest_image_filenames = glob(join(DATASET_FOLDER, , , , ))\n\nids = []\nrles = []\n filename  test_image_filenames:\n    label = filename.split()[]\n    slice_number = basename(filename).split()[]\n    ids.append()\n    predicted_rle = \n    rles.append(predicted_rle)\n\ndf = pd.DataFrame({: ids, : rles}).sort_values(by=[])\ndf.to_csv(, index=)\n</code></pre>",
      "votes": null,
      "replies": []
    }
  ],
  "raw_markdown_by_id": {
    "2530128": "Hi,\n\nJust looking at the data provided.\n\nKidney 5 and 6 images look exactly the same. Is this meant to be like that?",
    "2530134": "https://www.kaggle.com/perdigao1/kidney5-kidney6",
    "2530549": "The data in the test folder are just dummy values. You should only care about how the test data is structured.\nWhen reading the test data during inference, make sure that it is a generic way to extract all the TIFF files with the given test folder structure. Example:\n```python\nimport numpy as np # linear algebra\nimport pandas as pd # data processing, CSV file I/O (e.g. pd.read_csv)\nfrom glob import glob\nfrom os.path import basename, join\n\nDATASET_FOLDER = \"/kaggle/input/blood-vessel-segmentation\"\ntest_image_filenames = glob(join(DATASET_FOLDER, \"test\", \"*\", \"*\", \"*.tif\"))\n\nids = []\nrles = []\nfor filename in test_image_filenames:\n    label = filename.split('/')[5]\n    slice_number = basename(filename).split('.')[0]\n    ids.append(f\"{label}_{slice_number}\")\n    predicted_rle = # Your inference code\n    rles.append(predicted_rle)\n\ndf = pd.DataFrame({'id': ids, 'rle': rles}).sort_values(by=['id'])\ndf.to_csv('submission.csv', index=False)\n```"
  },
  "source": "meta"
}