{
  "id": 116012,
  "title": "Interpolating missing weather data",
  "url": "/competitions/ashrae-energy-prediction/discussion/116012",
  "author_name": "Henrique Mendonça",
  "post_date": "2019-11-06T15:31:24.164000",
  "votes": 11,
  "comment_count": 4,
  "views": 0,
  "content": "<p>Some people said that they were using interpolation to fill missing weather data points, but I wonder, what method performed best?</p>\n\n<p>I've tried a some different interpolation strategies. Cubic interpolation is generally the best option but the data is a bit too noisy for it, so 50/50% mix with linear interpolation seems better:\n<img src=\"https://www.googleapis.com/download/storage/v1/b/kaggle-user-content/o/inbox%2F451025%2F4e51a92c655d4ba4a6a38d7ac7e0a496%2Fdew_interpol.png?generation=1573054012540817&amp;alt=media\" alt=\"Interpolated dew temp\">\nNote: The blue is the missing data, now refilled!</p>\n\n<p>Edit: code (and data) are now available in <a href=\"https://www.kaggle.com/hmendonca/clean-weather-data-eda\">this kernel</a></p>",
  "messages": [
    {
      "id": 666885,
      "postDate": "2019-11-06T15:31:24.163Z",
      "content": "<p>Some people said that they were using interpolation to fill missing weather data points, but I wonder, what method performed best?</p>\n\n<p>I've tried a some different interpolation strategies. Cubic interpolation is generally the best option but the data is a bit too noisy for it, so 50/50% mix with linear interpolation seems better:\n<img src=\"https://www.googleapis.com/download/storage/v1/b/kaggle-user-content/o/inbox%2F451025%2F4e51a92c655d4ba4a6a38d7ac7e0a496%2Fdew_interpol.png?generation=1573054012540817&amp;alt=media\" alt=\"Interpolated dew temp\">\nNote: The blue is the missing data, now refilled!</p>\n\n<p>Edit: code (and data) are now available in <a href=\"https://www.kaggle.com/hmendonca/clean-weather-data-eda\">this kernel</a></p>",
      "rawMarkdown": "Some people said that they were using interpolation to fill missing weather data points, but I wonder, what method performed best?\n\nI've tried a some different interpolation strategies. Cubic interpolation is generally the best option but the data is a bit too noisy for it, so 50/50% mix with linear interpolation seems better:\n![Interpolated dew temp](https://www.googleapis.com/download/storage/v1/b/kaggle-user-content/o/inbox%2F451025%2F4e51a92c655d4ba4a6a38d7ac7e0a496%2Fdew_interpol.png?generation=1573054012540817&amp;alt=media)\nNote: The blue is the missing data, now refilled!\n\nEdit: code (and data) are now available in [this kernel](https://www.kaggle.com/hmendonca/clean-weather-data-eda)",
      "votes": 11
    },
    {
      "id": 666915,
      "postDate": "2019-11-06T16:13:00.707Z",
      "content": "<p>The code and cleaned data can be found in this kernel <a href=\"https://www.kaggle.com/hmendonca/clean-weather-data-eda\">https://www.kaggle.com/hmendonca/clean-weather-data-eda</a></p>",
      "rawMarkdown": "The code and cleaned data can be found in this kernel https://www.kaggle.com/hmendonca/clean-weather-data-eda",
      "votes": 2,
      "replies": [
        {
          "id": 667169,
          "postDate": "2019-11-06T22:33:08.990Z",
          "content": "<p>damn, now it's public! sorry :)</p>",
          "rawMarkdown": "damn, now it's public! sorry :)",
          "votes": 2
        },
        {
          "id": 667255,
          "postDate": "2019-11-07T02:45:31.620Z",
          "content": "<p>Thanks!Have you tried using this method to fill the missing data and submit the result?</p>",
          "rawMarkdown": "Thanks!Have you tried using this method to fill the missing data and submit the result?",
          "votes": 1
        },
        {
          "id": 668339,
          "postDate": "2019-11-08T09:41:55.773Z",
          "rawMarkdown": "",
          "isDeleted": true
        }
      ]
    }
  ],
  "comments": [
    {
      "id": 666915,
      "author_name": "Henrique Mendonça",
      "author_url": "",
      "post_date": "2019-11-06T16:13:00.707000",
      "content": "<p>The code and cleaned data can be found in this kernel <a href=\"https://www.kaggle.com/hmendonca/clean-weather-data-eda\">https://www.kaggle.com/hmendonca/clean-weather-data-eda</a></p>",
      "votes": 2,
      "replies": [
        {
          "id": 667169,
          "author_name": "Henrique Mendonça",
          "author_url": "",
          "post_date": "2019-11-06T22:33:08.990000",
          "content": "<p>damn, now it's public! sorry :)</p>",
          "votes": 2,
          "replies": []
        },
        {
          "id": 667255,
          "author_name": "Dxx",
          "author_url": "",
          "post_date": "2019-11-07T02:45:31.620000",
          "content": "<p>Thanks!Have you tried using this method to fill the missing data and submit the result?</p>",
          "votes": 1,
          "replies": []
        },
        {
          "id": 668339,
          "author_name": "",
          "author_url": "",
          "post_date": "2019-11-08T09:41:55.773000",
          "content": "",
          "votes": 0,
          "replies": []
        }
      ]
    }
  ],
  "raw_markdown_by_id": {
    "666885": "Some people said that they were using interpolation to fill missing weather data points, but I wonder, what method performed best?\n\nI've tried a some different interpolation strategies. Cubic interpolation is generally the best option but the data is a bit too noisy for it, so 50/50% mix with linear interpolation seems better:\n![Interpolated dew temp](https://www.googleapis.com/download/storage/v1/b/kaggle-user-content/o/inbox%2F451025%2F4e51a92c655d4ba4a6a38d7ac7e0a496%2Fdew_interpol.png?generation=1573054012540817&amp;alt=media)\nNote: The blue is the missing data, now refilled!\n\nEdit: code (and data) are now available in [this kernel](https://www.kaggle.com/hmendonca/clean-weather-data-eda)",
    "666915": "The code and cleaned data can be found in this kernel https://www.kaggle.com/hmendonca/clean-weather-data-eda"
  }
}