{
  "id": 314049,
  "title": "Missing images",
  "url": "/competitions/sorghum-id-fgvc-9/discussion/314049",
  "author_name": "",
  "post_date": "2022-03-20T16:23:11.826898100Z",
  "votes": 1,
  "comment_count": 1,
  "views": 0,
  "content": "<p>442 images mentioned in the csv are missing in the train_images folder.<br>\nUse this code to identify those images….<br>\n<code>data = pd.read_csv('../input/sorghum-id-fgvc-9/train_cultivar_mapping.csv')\nimages = data['image'].values\nhead = '../input/sorghum-id-fgvc-9/train_images/'\nnot_found = []\nfor im in tqdm(images):\n    if not os.path.exists(head + im):\n        not_found.append(im)\nprint(not_found)</code></p>",
  "messages": [
    {
      "id": "1729864",
      "postDate": "03/20/2022 16:23:11",
      "content": "<p>442 images mentioned in the csv are missing in the train_images folder.<br>\nUse this code to identify those images….<br>\n<code>data = pd.read_csv('../input/sorghum-id-fgvc-9/train_cultivar_mapping.csv')\nimages = data['image'].values\nhead = '../input/sorghum-id-fgvc-9/train_images/'\nnot_found = []\nfor im in tqdm(images):\n    if not os.path.exists(head + im):\n        not_found.append(im)\nprint(not_found)</code></p>",
      "rawMarkdown": "442 images mentioned in the csv are missing in the train_images folder.\nUse this code to identify those images....\n`data = pd.read_csv('../input/sorghum-id-fgvc-9/train_cultivar_mapping.csv')\nimages = data['image'].values\nhead = '../input/sorghum-id-fgvc-9/train_images/'\nnot_found = []\nfor im in tqdm(images):\n    if not os.path.exists(head + im):\n        not_found.append(im)\nprint(not_found)`",
      "votes": null
    },
    {
      "id": "1733720",
      "postDate": "03/24/2022 14:50:36",
      "content": "<p>Hi! Apologies for this -- just prior to launch we did a bit of data cleaning to remove some images that were super dark or super bright from both the training and test sets. It looks like the test set CSVs got updated but not the training set ones! I've sent over the updated one to the Kaggle team to hopefully get the file updated ASAP. In the meantime, I really appreciate you sharing your code to help others deal with these missing files!</p>",
      "rawMarkdown": "Hi! Apologies for this -- just prior to launch we did a bit of data cleaning to remove some images that were super dark or super bright from both the training and test sets. It looks like the test set CSVs got updated but not the training set ones! I've sent over the updated one to the Kaggle team to hopefully get the file updated ASAP. In the meantime, I really appreciate you sharing your code to help others deal with these missing files!",
      "votes": null
    }
  ],
  "comments": [
    {
      "id": 1733720,
      "author_name": "abbystylianou",
      "author_url": "",
      "post_date": "03/24/2022 14:50:36",
      "content": "<p>Hi! Apologies for this -- just prior to launch we did a bit of data cleaning to remove some images that were super dark or super bright from both the training and test sets. It looks like the test set CSVs got updated but not the training set ones! I've sent over the updated one to the Kaggle team to hopefully get the file updated ASAP. In the meantime, I really appreciate you sharing your code to help others deal with these missing files!</p>",
      "votes": null,
      "replies": []
    }
  ],
  "raw_markdown_by_id": {
    "1729864": "442 images mentioned in the csv are missing in the train_images folder.\nUse this code to identify those images....\n`data = pd.read_csv('../input/sorghum-id-fgvc-9/train_cultivar_mapping.csv')\nimages = data['image'].values\nhead = '../input/sorghum-id-fgvc-9/train_images/'\nnot_found = []\nfor im in tqdm(images):\n    if not os.path.exists(head + im):\n        not_found.append(im)\nprint(not_found)`",
    "1733720": "Hi! Apologies for this -- just prior to launch we did a bit of data cleaning to remove some images that were super dark or super bright from both the training and test sets. It looks like the test set CSVs got updated but not the training set ones! I've sent over the updated one to the Kaggle team to hopefully get the file updated ASAP. In the meantime, I really appreciate you sharing your code to help others deal with these missing files!"
  },
  "source": "meta"
}