{
  "id": 614083,
  "title": "Are There Any New Species in the Test Dataset?",
  "url": "/competitions/csiro-biomass/discussion/614083",
  "author_name": "",
  "post_date": "2025-11-01T00:52:21.835642500Z",
  "votes": 6,
  "comment_count": 5,
  "views": 0,
  "content": "<p>Hi. I have noticed that train.csv has a column \"Species\". I'm wondering if there are any new species in the test set photos that are not listed in train.csv? Thanks.</p>",
  "messages": [
    {
      "id": "3309551",
      "postDate": "11/01/2025 00:52:21",
      "content": "<p>Hi. I have noticed that train.csv has a column \"Species\". I'm wondering if there are any new species in the test set photos that are not listed in train.csv? Thanks.</p>",
      "rawMarkdown": "Hi. I have noticed that train.csv has a column \"Species\". I'm wondering if there are any new species in the test set photos that are not listed in train.csv? Thanks.",
      "votes": null
    },
    {
      "id": "3309620",
      "postDate": "11/01/2025 04:06:38",
      "content": "<p>Great question. The direct answer is that all the dominant species in the test set are the same as those in the training set.</p>\n<p>Please note that the \"Species\" column refers to the visually annotated dominant species in the images and <strong>does not include every species present</strong> (which would be practically impossible). Additionally, the \"Species\" column is only provided for the training set and is not available for the testing set (in case you didn't know it).</p>",
      "rawMarkdown": "Great question. The direct answer is that all the dominant species in the test set are the same as those in the training set.\n\nPlease note that the \"Species\" column refers to the visually annotated dominant species in the images and **does not include every species present** (which would be practically impossible). Additionally, the \"Species\" column is only provided for the training set and is not available for the testing set (in case you didn't know it).",
      "votes": null
    },
    {
      "id": "3315562",
      "postDate": "11/10/2025 16:59:29",
      "content": "<p>thank you for the response, just noticed that the test data doesn't contain any features that exist in the train data like state , species , height and … \nso my question is why you gave such info in train data or why didn't you give these features in test ?  :)\nand apparently we should only predict the target by image analysis , right? </p>",
      "rawMarkdown": "thank you for the response, just noticed that the test data doesn't contain any features that exist in the train data like state , species , height and ... \nso my question is why you gave such info in train data or why didn't you give these features in test ?  :)\nand apparently we should only predict the target by image analysis , right?",
      "votes": null
    },
    {
      "id": "3315563",
      "postDate": "11/10/2025 17:00:27",
      "content": "<p>You can build decent models that predict those features - might be a path to better LB score.</p>",
      "rawMarkdown": "You can build decent models that predict those features - might be a path to better LB score.",
      "votes": null
    },
    {
      "id": "3315595",
      "postDate": "11/10/2025 17:16:00",
      "content": "<p>thank you … yes, I was thinking about that but wanted to make sure that does the private test data contains those features or has only the image file?</p>",
      "rawMarkdown": "thank you ... yes, I was thinking about that but wanted to make sure that does the private test data contains those features or has only the image file?",
      "votes": null
    },
    {
      "id": "3315739",
      "postDate": "11/10/2025 18:38:01",
      "content": "<p>Test only includes image files.  </p>",
      "rawMarkdown": "Test only includes image files.",
      "votes": null
    }
  ],
  "comments": [
    {
      "id": 3309620,
      "author_name": "qiyuliaocsiro",
      "author_url": "",
      "post_date": "11/01/2025 04:06:38",
      "content": "<p>Great question. The direct answer is that all the dominant species in the test set are the same as those in the training set.</p>\n<p>Please note that the \"Species\" column refers to the visually annotated dominant species in the images and <strong>does not include every species present</strong> (which would be practically impossible). Additionally, the \"Species\" column is only provided for the training set and is not available for the testing set (in case you didn't know it).</p>",
      "votes": null,
      "replies": [
        {
          "id": 3315562,
          "author_name": "malihealavi",
          "author_url": "",
          "post_date": "11/10/2025 16:59:29",
          "content": "<p>thank you for the response, just noticed that the test data doesn't contain any features that exist in the train data like state , species , height and … \nso my question is why you gave such info in train data or why didn't you give these features in test ?  :)\nand apparently we should only predict the target by image analysis , right? </p>",
          "votes": null,
          "replies": [
            {
              "id": 3315563,
              "author_name": "pcjimmmy",
              "author_url": "",
              "post_date": "11/10/2025 17:00:27",
              "content": "<p>You can build decent models that predict those features - might be a path to better LB score.</p>",
              "votes": null,
              "replies": [
                {
                  "id": 3315595,
                  "author_name": "malihealavi",
                  "author_url": "",
                  "post_date": "11/10/2025 17:16:00",
                  "content": "<p>thank you … yes, I was thinking about that but wanted to make sure that does the private test data contains those features or has only the image file?</p>",
                  "votes": null,
                  "replies": [
                    {
                      "id": 3315739,
                      "author_name": "pcjimmmy",
                      "author_url": "",
                      "post_date": "11/10/2025 18:38:01",
                      "content": "<p>Test only includes image files.  </p>",
                      "votes": null,
                      "replies": []
                    }
                  ]
                }
              ]
            }
          ]
        }
      ]
    }
  ],
  "raw_markdown_by_id": {
    "3309551": "Hi. I have noticed that train.csv has a column \"Species\". I'm wondering if there are any new species in the test set photos that are not listed in train.csv? Thanks.",
    "3309620": "Great question. The direct answer is that all the dominant species in the test set are the same as those in the training set.\n\nPlease note that the \"Species\" column refers to the visually annotated dominant species in the images and **does not include every species present** (which would be practically impossible). Additionally, the \"Species\" column is only provided for the training set and is not available for the testing set (in case you didn't know it).",
    "3315562": "thank you for the response, just noticed that the test data doesn't contain any features that exist in the train data like state , species , height and ... \nso my question is why you gave such info in train data or why didn't you give these features in test ?  :)\nand apparently we should only predict the target by image analysis , right?",
    "3315563": "You can build decent models that predict those features - might be a path to better LB score.",
    "3315595": "thank you ... yes, I was thinking about that but wanted to make sure that does the private test data contains those features or has only the image file?",
    "3315739": "Test only includes image files."
  },
  "source": "meta"
}