{
  "id": 337731,
  "title": "Missing Value Imputation",
  "url": "/competitions/amex-default-prediction/discussion/337731",
  "author_name": "",
  "post_date": "2022-07-17T10:51:42.449474Z",
  "votes": 1,
  "comment_count": 1,
  "views": 0,
  "content": "<p>While doing missing value imputation, sklearn KNNImputer is dropping sample rows in which the majority of the feature columns are missing. There is no mention of this behavior in the <a href=\"https://scikit-learn.org/stable/modules/generated/sklearn.impute.KNNImputer.html\" target=\"_blank\">sklearn documentation</a>.</p>\n<p>Is this an expected behavior from KNNImputer?</p>",
  "messages": [
    {
      "id": "1859015",
      "postDate": "07/17/2022 10:51:42",
      "content": "<p>While doing missing value imputation, sklearn KNNImputer is dropping sample rows in which the majority of the feature columns are missing. There is no mention of this behavior in the <a href=\"https://scikit-learn.org/stable/modules/generated/sklearn.impute.KNNImputer.html\" target=\"_blank\">sklearn documentation</a>.</p>\n<p>Is this an expected behavior from KNNImputer?</p>",
      "rawMarkdown": "While doing missing value imputation, sklearn KNNImputer is dropping sample rows in which the majority of the feature columns are missing. There is no mention of this behavior in the [sklearn documentation](https://scikit-learn.org/stable/modules/generated/sklearn.impute.KNNImputer.html).\n\nIs this an expected behavior from KNNImputer?",
      "votes": null
    },
    {
      "id": "1859497",
      "postDate": "07/17/2022 17:20:08",
      "content": "<p>Deep dived further, it was not expected. There was an issue with df merge.</p>",
      "rawMarkdown": "Deep dived further, it was not expected. There was an issue with df merge.",
      "votes": null
    }
  ],
  "comments": [
    {
      "id": 1859497,
      "author_name": "crashoverdrive",
      "author_url": "",
      "post_date": "07/17/2022 17:20:08",
      "content": "<p>Deep dived further, it was not expected. There was an issue with df merge.</p>",
      "votes": null,
      "replies": []
    }
  ],
  "raw_markdown_by_id": {
    "1859015": "While doing missing value imputation, sklearn KNNImputer is dropping sample rows in which the majority of the feature columns are missing. There is no mention of this behavior in the [sklearn documentation](https://scikit-learn.org/stable/modules/generated/sklearn.impute.KNNImputer.html).\n\nIs this an expected behavior from KNNImputer?",
    "1859497": "Deep dived further, it was not expected. There was an issue with df merge."
  },
  "source": "meta"
}