{
  "id": 499737,
  "title": "How submission!! ",
  "url": "/competitions/leap-atmospheric-physics-ai-climsim/discussion/499737",
  "author_name": "Michael Kamal",
  "post_date": "2024-05-02T19:54:48.475000",
  "votes": -1,
  "comment_count": 7,
  "views": 0,
  "content": "<p>I can't understand what i should be do exactly<br>\nAfter predict<br>\nIs<br>\n submission directly or<br>\nPredict * weight or<br>\nAnother something<br>\nPlease help me, THANKS</p>",
  "messages": [
    {
      "id": 2789931,
      "postDate": "2024-05-02T22:56:52.870Z",
      "content": "<p>Hi, if you look at the first row of the sample submission file (every row is the same), it has a number for each column.  That number is the column weight.  Multiply your predictions by the column weight, and that new number is what you should save to your submission file.  Does that make sense?</p>",
      "rawMarkdown": "Hi, if you look at the first row of the sample submission file (every row is the same), it has a number for each column.  That number is the column weight.  Multiply your predictions by the column weight, and that new number is what you should save to your submission file.  Does that make sense?",
      "votes": 1,
      "replies": [
        {
          "id": 2789948,
          "postDate": "2024-05-02T23:33:15.150Z",
          "content": "<p>Yes, thanks for you answer. Now if i get score with negative, this mean model is bad right? </p>",
          "rawMarkdown": "Yes, thanks for you answer. Now if i get score with negative, this mean model is bad right? ",
          "replies": [
            {
              "id": 2790019,
              "postDate": "2024-05-03T01:14:53.177Z",
              "content": "<p>Yes, that's right.  You'll get a zero if you guess the mean for each column, so if you're weighting correctly it takes a pretty bad model to go below zero.</p>",
              "rawMarkdown": "Yes, that's right.  You'll get a zero if you guess the mean for each column, so if you're weighting correctly it takes a pretty bad model to go below zero."
            },
            {
              "id": 2790505,
              "postDate": "2024-05-03T08:01:30.230Z",
              "content": "<p>Thank you very much </p>",
              "rawMarkdown": "Thank you very much "
            },
            {
              "id": 2795744,
              "postDate": "2024-05-05T23:17:06.873Z",
              "content": "<p>Hello again, i want more sure this way is right, Predict * weight or<br>\nAnother something</p>",
              "rawMarkdown": "Hello again, i want more sure this way is right, Predict * weight or\nAnother something"
            },
            {
              "id": 2795833,
              "postDate": "2024-05-06T02:37:15.940Z",
              "content": "<p>Yes, for each column multiply that column's prediction times that column's weight.</p>",
              "rawMarkdown": "Yes, for each column multiply that column's prediction times that column's weight."
            },
            {
              "id": 2795883,
              "postDate": "2024-05-06T03:23:26.367Z",
              "content": "<p>Thank you for help me 🥰</p>",
              "rawMarkdown": "Thank you for help me 🥰"
            }
          ]
        }
      ]
    },
    {
      "id": 2789756,
      "postDate": "2024-05-02T19:54:48.477Z",
      "content": "<p>I can't understand what i should be do exactly<br>\nAfter predict<br>\nIs<br>\n submission directly or<br>\nPredict * weight or<br>\nAnother something<br>\nPlease help me, THANKS</p>",
      "rawMarkdown": "I can't understand what i should be do exactly\nAfter predict\nIs\n submission directly or\nPredict * weight or\nAnother something\nPlease help me, THANKS",
      "votes": -1
    }
  ],
  "comments": [
    {
      "id": 2789931,
      "author_name": "Ted K",
      "author_url": "",
      "post_date": "2024-05-02T22:56:52.870000",
      "content": "<p>Hi, if you look at the first row of the sample submission file (every row is the same), it has a number for each column.  That number is the column weight.  Multiply your predictions by the column weight, and that new number is what you should save to your submission file.  Does that make sense?</p>",
      "votes": 1,
      "replies": [
        {
          "id": 2789948,
          "author_name": "Michael Kamal",
          "author_url": "",
          "post_date": "2024-05-02T23:33:15.150000",
          "content": "<p>Yes, thanks for you answer. Now if i get score with negative, this mean model is bad right? </p>",
          "votes": 0,
          "replies": [
            {
              "id": 2790019,
              "author_name": "Ted K",
              "author_url": "",
              "post_date": "2024-05-03T01:14:53.177000",
              "content": "<p>Yes, that's right.  You'll get a zero if you guess the mean for each column, so if you're weighting correctly it takes a pretty bad model to go below zero.</p>",
              "votes": 0,
              "replies": []
            },
            {
              "id": 2790505,
              "author_name": "Michael Kamal",
              "author_url": "",
              "post_date": "2024-05-03T08:01:30.230000",
              "content": "<p>Thank you very much </p>",
              "votes": 0,
              "replies": []
            },
            {
              "id": 2795744,
              "author_name": "Michael Kamal",
              "author_url": "",
              "post_date": "2024-05-05T23:17:06.873000",
              "content": "<p>Hello again, i want more sure this way is right, Predict * weight or<br>\nAnother something</p>",
              "votes": 0,
              "replies": []
            },
            {
              "id": 2795833,
              "author_name": "Ted K",
              "author_url": "",
              "post_date": "2024-05-06T02:37:15.940000",
              "content": "<p>Yes, for each column multiply that column's prediction times that column's weight.</p>",
              "votes": 0,
              "replies": []
            },
            {
              "id": 2795883,
              "author_name": "Michael Kamal",
              "author_url": "",
              "post_date": "2024-05-06T03:23:26.367000",
              "content": "<p>Thank you for help me 🥰</p>",
              "votes": 0,
              "replies": []
            }
          ]
        }
      ]
    }
  ],
  "raw_markdown_by_id": {
    "2789931": "Hi, if you look at the first row of the sample submission file (every row is the same), it has a number for each column.  That number is the column weight.  Multiply your predictions by the column weight, and that new number is what you should save to your submission file.  Does that make sense?",
    "2789756": "I can't understand what i should be do exactly\nAfter predict\nIs\n submission directly or\nPredict * weight or\nAnother something\nPlease help me, THANKS"
  }
}