{
  "id": 212306,
  "title": "Notebook running",
  "url": "/competitions/hubmap-kidney-segmentation/discussion/212306",
  "author_name": "",
  "post_date": "2021-01-18T13:37:37.154838800Z",
  "votes": 1,
  "comment_count": 3,
  "views": 0,
  "content": "<p>I have submit my the notebook a few hours before, but the status is not changed. Is it a code bug? anyone can help ?<br>\n<img src=\"https://www.googleapis.com/download/storage/v1/b/kaggle-forum-message-attachments/o/inbox%2F1069548%2Fe92ab6f13f2d87658b707e20de3d74f7%2Fscore.png?generation=1610977030974201&amp;alt=media\" alt=\"\"></p>",
  "messages": [
    {
      "id": "1158252",
      "postDate": "01/18/2021 13:37:37",
      "content": "<p>I have submit my the notebook a few hours before, but the status is not changed. Is it a code bug? anyone can help ?<br>\n<img src=\"https://www.googleapis.com/download/storage/v1/b/kaggle-forum-message-attachments/o/inbox%2F1069548%2Fe92ab6f13f2d87658b707e20de3d74f7%2Fscore.png?generation=1610977030974201&amp;alt=media\" alt=\"\"></p>",
      "rawMarkdown": "I have submit my the notebook a few hours before, but the status is not changed. Is it a code bug? anyone can help ?\n![](https://www.googleapis.com/download/storage/v1/b/kaggle-forum-message-attachments/o/inbox%2F1069548%2Fe92ab6f13f2d87658b707e20de3d74f7%2Fscore.png?generation=1610977030974201&alt=media)",
      "votes": null
    },
    {
      "id": "1158562",
      "postDate": "01/18/2021 16:41:11",
      "content": "<p>How much time for you to run and save the document?</p>\n<p>if its more then 2 hours the scoring will take time</p>",
      "rawMarkdown": "How much time for you to run and save the document?\n\nif its more then 2 hours the scoring will take time",
      "votes": null
    },
    {
      "id": "1159314",
      "postDate": "01/19/2021 07:29:30",
      "content": "<p>There is probably an image in the private test set that is bigger than any image in the public test set. If committing your kernel took much less time, it probably means that it is using too much RAM because of that big image in the private test set. In that case, the submission keeps running until it times out.</p>\n<p>You can load an image efficiently by using NumPy memmap objects instead of arrays. That can be done as follows.</p>\n<pre><code>import tifffile as tiff\n\ndef load_image(filename, verbose=False):\n    img_id = filename.split(\"/\")[-1].split(\".\")[0]\n    with tiff.TiffFile(filename) as f:\n        img = f.pages[0].asarray(\"%s.npy\" %(img_id,))\n    if img.shape[0] == 3:\n        img = img.transpose((1,2,0))\n    if verbose:\n        print(\"Initial size:\", img.shape)\n    return img\n</code></pre>",
      "rawMarkdown": "There is probably an image in the private test set that is bigger than any image in the public test set. If committing your kernel took much less time, it probably means that it is using too much RAM because of that big image in the private test set. In that case, the submission keeps running until it times out.\n\nYou can load an image efficiently by using NumPy memmap objects instead of arrays. That can be done as follows.\n\n```\nimport tifffile as tiff\n\ndef load_image(filename, verbose=False):\n    img_id = filename.split(\"/\")[-1].split(\".\")[0]\n    with tiff.TiffFile(filename) as f:\n        img = f.pages[0].asarray(\"%s.npy\" %(img_id,))\n    if img.shape[0] == 3:\n        img = img.transpose((1,2,0))\n    if verbose:\n        print(\"Initial size:\", img.shape)\n    return img\n```",
      "votes": null
    },
    {
      "id": "1159654",
      "postDate": "01/19/2021 11:37:48",
      "content": "<p>I modified the way of reading the tiff image, and it can be submitted successfully. I guess it is because of insufficient memory space.</p>",
      "rawMarkdown": "I modified the way of reading the tiff image, and it can be submitted successfully. I guess it is because of insufficient memory space.",
      "votes": null
    }
  ],
  "comments": [
    {
      "id": 1158562,
      "author_name": "kingabzpro",
      "author_url": "",
      "post_date": "01/18/2021 16:41:11",
      "content": "<p>How much time for you to run and save the document?</p>\n<p>if its more then 2 hours the scoring will take time</p>",
      "votes": null,
      "replies": [
        {
          "id": 1159654,
          "author_name": "davinca",
          "author_url": "",
          "post_date": "01/19/2021 11:37:48",
          "content": "<p>I modified the way of reading the tiff image, and it can be submitted successfully. I guess it is because of insufficient memory space.</p>",
          "votes": null,
          "replies": []
        }
      ]
    },
    {
      "id": 1159314,
      "author_name": "dannyelkallab",
      "author_url": "",
      "post_date": "01/19/2021 07:29:30",
      "content": "<p>There is probably an image in the private test set that is bigger than any image in the public test set. If committing your kernel took much less time, it probably means that it is using too much RAM because of that big image in the private test set. In that case, the submission keeps running until it times out.</p>\n<p>You can load an image efficiently by using NumPy memmap objects instead of arrays. That can be done as follows.</p>\n<pre><code>import tifffile as tiff\n\ndef load_image(filename, verbose=False):\n    img_id = filename.split(\"/\")[-1].split(\".\")[0]\n    with tiff.TiffFile(filename) as f:\n        img = f.pages[0].asarray(\"%s.npy\" %(img_id,))\n    if img.shape[0] == 3:\n        img = img.transpose((1,2,0))\n    if verbose:\n        print(\"Initial size:\", img.shape)\n    return img\n</code></pre>",
      "votes": null,
      "replies": []
    }
  ],
  "raw_markdown_by_id": {
    "1158252": "I have submit my the notebook a few hours before, but the status is not changed. Is it a code bug? anyone can help ?\n![](https://www.googleapis.com/download/storage/v1/b/kaggle-forum-message-attachments/o/inbox%2F1069548%2Fe92ab6f13f2d87658b707e20de3d74f7%2Fscore.png?generation=1610977030974201&alt=media)",
    "1158562": "How much time for you to run and save the document?\n\nif its more then 2 hours the scoring will take time",
    "1159314": "There is probably an image in the private test set that is bigger than any image in the public test set. If committing your kernel took much less time, it probably means that it is using too much RAM because of that big image in the private test set. In that case, the submission keeps running until it times out.\n\nYou can load an image efficiently by using NumPy memmap objects instead of arrays. That can be done as follows.\n\n```\nimport tifffile as tiff\n\ndef load_image(filename, verbose=False):\n    img_id = filename.split(\"/\")[-1].split(\".\")[0]\n    with tiff.TiffFile(filename) as f:\n        img = f.pages[0].asarray(\"%s.npy\" %(img_id,))\n    if img.shape[0] == 3:\n        img = img.transpose((1,2,0))\n    if verbose:\n        print(\"Initial size:\", img.shape)\n    return img\n```",
    "1159654": "I modified the way of reading the tiff image, and it can be submitted successfully. I guess it is because of insufficient memory space."
  },
  "source": "meta"
}