{
  "id": 94284,
  "title": "Atom index and Atom index0 meaning",
  "url": "/competitions/champs-scalar-coupling/discussion/94284",
  "author_name": "",
  "post_date": "2019-06-03T18:46:17.599045500Z",
  "votes": 1,
  "comment_count": 4,
  "views": 0,
  "content": "<p>In <strong>mulliken_charges.csv</strong> *atom_index* column is provided and in train_scalar_coupling.csv two columns with name <strong>atom_index_0</strong> and &gt; atom_index_1 are available. I know that <strong>scalar coupling constant</strong> is there for *atom_index_0* and *atom_index_1*. </p>\n\n<p>Do the atom_index_0 and atom_index are same? In train.csv for which index I have to take the mullikan  charges?</p>",
  "messages": [
    {
      "id": "542359",
      "postDate": "06/03/2019 18:46:17",
      "content": "<p>In <strong>mulliken_charges.csv</strong> *atom_index* column is provided and in train_scalar_coupling.csv two columns with name <strong>atom_index_0</strong> and &gt; atom_index_1 are available. I know that <strong>scalar coupling constant</strong> is there for *atom_index_0* and *atom_index_1*. </p>\n\n<p>Do the atom_index_0 and atom_index are same? In train.csv for which index I have to take the mullikan  charges?</p>",
      "rawMarkdown": "In **mulliken_charges.csv** *atom_index* column is provided and in train_scalar_coupling.csv two columns with name **atom_index_0** and &gt; atom_index_1 are available. I know that **scalar coupling constant** is there for *atom_index_0* and *atom_index_1*. \n\nDo the atom_index_0 and atom_index are same? In train.csv for which index I have to take the mullikan  charges?",
      "votes": null
    },
    {
      "id": "542387",
      "postDate": "06/03/2019 20:07:02",
      "content": "<p>In the training set, for each row you have two atoms, and the magnetic coupling between these two atoms. Since there are two atoms, you need a subscript on \"atom index\" to help you identify which atom you are dealing with. So \"atom index 0\" (resp. \"atom index 1\") corresponds to \"atom 0\" (resp. \"atom 1\") in the training set.</p>\n\n<p>In the mulliken charges file, there is only one atom on each row, and its associated mulliken charge. Therefore, you don't need a subscript and \"atom index\" corresponds to \"atom\".</p>",
      "rawMarkdown": "In the training set, for each row you have two atoms, and the magnetic coupling between these two atoms. Since there are two atoms, you need a subscript on \"atom index\" to help you identify which atom you are dealing with. So \"atom index 0\" (resp. \"atom index 1\") corresponds to \"atom 0\" (resp. \"atom 1\") in the training set.\n\nIn the mulliken charges file, there is only one atom on each row, and its associated mulliken charge. Therefore, you don't need a subscript and \"atom index\" corresponds to \"atom\".",
      "votes": null
    },
    {
      "id": "542769",
      "postDate": "06/04/2019 04:45:43",
      "content": "<p>Thnx <a href=\"/borisdee\">@borisdee</a> For train set only Mulliken charge is provided. What about the test set. DO we need to search for external data? (i.e .Mulliken charge of test data's)</p>",
      "rawMarkdown": "Thnx @borisdee For train set only Mulliken charge is provided. What about the test set. DO we need to search for external data? (i.e .Mulliken charge of test data's)",
      "votes": null
    },
    {
      "id": "542830",
      "postDate": "06/04/2019 06:06:48",
      "content": "<p>You won't find any external data related to the Mulliken charges of the molecules in the test set. My guess is that it is possible to predict the Mulliken charges of the test set using those of the train set.\nSome kernels have calculated averages of the mulliken charges and used this in the test set. I don't know what it's worth.</p>",
      "rawMarkdown": "You won't find any external data related to the Mulliken charges of the molecules in the test set. My guess is that it is possible to predict the Mulliken charges of the test set using those of the train set.\nSome kernels have calculated averages of the mulliken charges and used this in the test set. I don't know what it's worth.",
      "votes": null
    },
    {
      "id": "543768",
      "postDate": "06/04/2019 19:48:15",
      "content": "<p>For test we have only test and structures.csv . So i will try to find the mulliken from structures.csv if possible. It may provide some good insights</p>",
      "rawMarkdown": "For test we have only test and structures.csv . So i will try to find the mulliken from structures.csv if possible. It may provide some good insights",
      "votes": null
    }
  ],
  "comments": [
    {
      "id": 542387,
      "author_name": "borisdee",
      "author_url": "",
      "post_date": "06/03/2019 20:07:02",
      "content": "<p>In the training set, for each row you have two atoms, and the magnetic coupling between these two atoms. Since there are two atoms, you need a subscript on \"atom index\" to help you identify which atom you are dealing with. So \"atom index 0\" (resp. \"atom index 1\") corresponds to \"atom 0\" (resp. \"atom 1\") in the training set.</p>\n\n<p>In the mulliken charges file, there is only one atom on each row, and its associated mulliken charge. Therefore, you don't need a subscript and \"atom index\" corresponds to \"atom\".</p>",
      "votes": null,
      "replies": []
    },
    {
      "id": 542769,
      "author_name": "saravanansaminathan",
      "author_url": "",
      "post_date": "06/04/2019 04:45:43",
      "content": "<p>Thnx <a href=\"/borisdee\">@borisdee</a> For train set only Mulliken charge is provided. What about the test set. DO we need to search for external data? (i.e .Mulliken charge of test data's)</p>",
      "votes": null,
      "replies": [
        {
          "id": 542830,
          "author_name": "borisdee",
          "author_url": "",
          "post_date": "06/04/2019 06:06:48",
          "content": "<p>You won't find any external data related to the Mulliken charges of the molecules in the test set. My guess is that it is possible to predict the Mulliken charges of the test set using those of the train set.\nSome kernels have calculated averages of the mulliken charges and used this in the test set. I don't know what it's worth.</p>",
          "votes": null,
          "replies": []
        },
        {
          "id": 543768,
          "author_name": "saravanansaminathan",
          "author_url": "",
          "post_date": "06/04/2019 19:48:15",
          "content": "<p>For test we have only test and structures.csv . So i will try to find the mulliken from structures.csv if possible. It may provide some good insights</p>",
          "votes": null,
          "replies": []
        }
      ]
    }
  ],
  "raw_markdown_by_id": {
    "542359": "In **mulliken_charges.csv** *atom_index* column is provided and in train_scalar_coupling.csv two columns with name **atom_index_0** and &gt; atom_index_1 are available. I know that **scalar coupling constant** is there for *atom_index_0* and *atom_index_1*. \n\nDo the atom_index_0 and atom_index are same? In train.csv for which index I have to take the mullikan  charges?",
    "542387": "In the training set, for each row you have two atoms, and the magnetic coupling between these two atoms. Since there are two atoms, you need a subscript on \"atom index\" to help you identify which atom you are dealing with. So \"atom index 0\" (resp. \"atom index 1\") corresponds to \"atom 0\" (resp. \"atom 1\") in the training set.\n\nIn the mulliken charges file, there is only one atom on each row, and its associated mulliken charge. Therefore, you don't need a subscript and \"atom index\" corresponds to \"atom\".",
    "542769": "Thnx @borisdee For train set only Mulliken charge is provided. What about the test set. DO we need to search for external data? (i.e .Mulliken charge of test data's)",
    "542830": "You won't find any external data related to the Mulliken charges of the molecules in the test set. My guess is that it is possible to predict the Mulliken charges of the test set using those of the train set.\nSome kernels have calculated averages of the mulliken charges and used this in the test set. I don't know what it's worth.",
    "543768": "For test we have only test and structures.csv . So i will try to find the mulliken from structures.csv if possible. It may provide some good insights"
  },
  "source": "meta"
}