{
  "id": 104690,
  "title": "Density Functional Theory",
  "url": "/competitions/champs-scalar-coupling/discussion/104690",
  "author_name": "Scirpus",
  "post_date": "2019-08-18T13:20:19.705000",
  "votes": 1,
  "comment_count": 0,
  "views": 0,
  "content": "<p>Most modern ab-initio calculations use Density Functional Theory using VASP if you are rich or similar programs if you are poor.</p>\n\n<p>What gets me is that few people in the kernels (sorry notebooks!) are using the charge information.  Have a look at edges in train that have similar atoms, charges and distances - to a first approximation they have similar coupling constants - so if you aren't using them you are nuts!</p>\n\n<p>Try at the very least to predict the charges - they might not be identical to the actual ones but similar H charge, Other Charge, Bond Distance, Other AtomType  quadruplets may allow you to use a KNN or similar to do some clever averaging. </p>",
  "messages": [
    {
      "id": 602026,
      "postDate": "2019-08-18T13:20:19.707Z",
      "content": "<p>Most modern ab-initio calculations use Density Functional Theory using VASP if you are rich or similar programs if you are poor.</p>\n\n<p>What gets me is that few people in the kernels (sorry notebooks!) are using the charge information.  Have a look at edges in train that have similar atoms, charges and distances - to a first approximation they have similar coupling constants - so if you aren't using them you are nuts!</p>\n\n<p>Try at the very least to predict the charges - they might not be identical to the actual ones but similar H charge, Other Charge, Bond Distance, Other AtomType  quadruplets may allow you to use a KNN or similar to do some clever averaging. </p>",
      "rawMarkdown": "Most modern ab-initio calculations use Density Functional Theory using VASP if you are rich or similar programs if you are poor.\n\nWhat gets me is that few people in the kernels (sorry notebooks!) are using the charge information.  Have a look at edges in train that have similar atoms, charges and distances - to a first approximation they have similar coupling constants - so if you aren't using them you are nuts!\n\nTry at the very least to predict the charges - they might not be identical to the actual ones but similar H charge, Other Charge, Bond Distance, Other AtomType  quadruplets may allow you to use a KNN or similar to do some clever averaging. ",
      "votes": 1
    }
  ],
  "comments": [],
  "raw_markdown_by_id": {
    "602026": "Most modern ab-initio calculations use Density Functional Theory using VASP if you are rich or similar programs if you are poor.\n\nWhat gets me is that few people in the kernels (sorry notebooks!) are using the charge information.  Have a look at edges in train that have similar atoms, charges and distances - to a first approximation they have similar coupling constants - so if you aren't using them you are nuts!\n\nTry at the very least to predict the charges - they might not be identical to the actual ones but similar H charge, Other Charge, Bond Distance, Other AtomType  quadruplets may allow you to use a KNN or similar to do some clever averaging. "
  }
}