{
  "id": 363834,
  "title": "How is session defined?",
  "url": "/competitions/otto-recommender-system/discussion/363834",
  "author_name": "",
  "post_date": "2022-11-03T10:13:49.430978700Z",
  "votes": 7,
  "comment_count": 1,
  "views": 0,
  "content": "<p>How is a session defined in the dataset? I see that some sessions last quite long (2022-07-31 to 2022-08-21). Is it a data preprocessing issue fro my side or are the sessions really this long? </p>\n<p>This is how I convert <code>ts</code> into <code>date_time</code>.</p>\n<pre><code>train[\"date_time\"] = pd.to_datetime(train.ts, unit=\"ms\")\n</code></pre>",
  "messages": [
    {
      "id": "2015492",
      "postDate": "11/03/2022 10:13:49",
      "content": "<p>How is a session defined in the dataset? I see that some sessions last quite long (2022-07-31 to 2022-08-21). Is it a data preprocessing issue fro my side or are the sessions really this long? </p>\n<p>This is how I convert <code>ts</code> into <code>date_time</code>.</p>\n<pre><code>train[\"date_time\"] = pd.to_datetime(train.ts, unit=\"ms\")\n</code></pre>",
      "rawMarkdown": "How is a session defined in the dataset? I see that some sessions last quite long (2022-07-31 to 2022-08-21). Is it a data preprocessing issue fro my side or are the sessions really this long? \n\nThis is how I convert `ts` into `date_time`.\n\n```\ntrain[\"date_time\"] = pd.to_datetime(train.ts, unit=\"ms\")\n```",
      "votes": null
    },
    {
      "id": "2015500",
      "postDate": "11/03/2022 10:27:43",
      "content": "<p>Hey <a href=\"https://www.kaggle.com/tarratana\" target=\"_blank\">@tarratana</a>, my comment in the related <a href=\"https://www.kaggle.com/competitions/otto-recommender-system/discussion/363554#2015486\" target=\"_blank\">discussion</a> should answer your question.<br>\n<strong>TLDR</strong>: Your preprocessing looks fine, and some sessions are actually that long.</p>",
      "rawMarkdown": "Hey @tarratana, my comment in the related [discussion](https://www.kaggle.com/competitions/otto-recommender-system/discussion/363554#2015486) should answer your question.\n**TLDR**: Your preprocessing looks fine, and some sessions are actually that long.",
      "votes": null
    }
  ],
  "comments": [
    {
      "id": 2015500,
      "author_name": "pnormann",
      "author_url": "",
      "post_date": "11/03/2022 10:27:43",
      "content": "<p>Hey <a href=\"https://www.kaggle.com/tarratana\" target=\"_blank\">@tarratana</a>, my comment in the related <a href=\"https://www.kaggle.com/competitions/otto-recommender-system/discussion/363554#2015486\" target=\"_blank\">discussion</a> should answer your question.<br>\n<strong>TLDR</strong>: Your preprocessing looks fine, and some sessions are actually that long.</p>",
      "votes": null,
      "replies": []
    }
  ],
  "raw_markdown_by_id": {
    "2015492": "How is a session defined in the dataset? I see that some sessions last quite long (2022-07-31 to 2022-08-21). Is it a data preprocessing issue fro my side or are the sessions really this long? \n\nThis is how I convert `ts` into `date_time`.\n\n```\ntrain[\"date_time\"] = pd.to_datetime(train.ts, unit=\"ms\")\n```",
    "2015500": "Hey @tarratana, my comment in the related [discussion](https://www.kaggle.com/competitions/otto-recommender-system/discussion/363554#2015486) should answer your question.\n**TLDR**: Your preprocessing looks fine, and some sessions are actually that long."
  },
  "source": "meta"
}