{
  "id": 66654,
  "title": "Can't find the target?",
  "url": "/competitions/ga-customer-revenue-prediction/discussion/66654",
  "author_name": "",
  "post_date": "2018-09-24T00:36:09.037997700Z",
  "votes": 5,
  "comment_count": 5,
  "views": 0,
  "content": "<p>Forgive me for being an idiot, but when I look at <code>train.csv</code> on <a href=\"https://www.kaggle.com/c/ga-customer-revenue-prediction/data\">https://www.kaggle.com/c/ga-customer-revenue-prediction/data</a> and it says:</p>\n\n<blockquote>\n  <p>There are multiple columns which contain JSON blobs of varying depth. In one of those JSON columns, totals, the sub-column transactionRevenue contains the revenue information we are trying to predict. This sub-column exists only for the training data.</p>\n</blockquote>\n\n<p>...but I don't see that subcolumn anywhere! &gt;.&lt;</p>\n\n<p>What am I missing?</p>",
  "messages": [
    {
      "id": "392545",
      "postDate": "09/24/2018 00:36:09",
      "content": "<p>Forgive me for being an idiot, but when I look at <code>train.csv</code> on <a href=\"https://www.kaggle.com/c/ga-customer-revenue-prediction/data\">https://www.kaggle.com/c/ga-customer-revenue-prediction/data</a> and it says:</p>\n\n<blockquote>\n  <p>There are multiple columns which contain JSON blobs of varying depth. In one of those JSON columns, totals, the sub-column transactionRevenue contains the revenue information we are trying to predict. This sub-column exists only for the training data.</p>\n</blockquote>\n\n<p>...but I don't see that subcolumn anywhere! &gt;.&lt;</p>\n\n<p>What am I missing?</p>",
      "rawMarkdown": "Forgive me for being an idiot, but when I look at `train.csv` on https://www.kaggle.com/c/ga-customer-revenue-prediction/data and it says:\n\n&gt; There are multiple columns which contain JSON blobs of varying depth. In one of those JSON columns, totals, the sub-column transactionRevenue contains the revenue information we are trying to predict. This sub-column exists only for the training data.\n\n...but I don't see that subcolumn anywhere! &gt;.&lt;\n\nWhat am I missing?",
      "votes": null
    },
    {
      "id": "392546",
      "postDate": "09/24/2018 00:39:44",
      "content": "<p><img src=\"https://i.imgur.com/u2N1Y8r.png\" alt=\"enter image description here\"></p>",
      "rawMarkdown": "![enter image description here][1]\n\n  [1]: https://i.imgur.com/u2N1Y8r.png",
      "votes": null
    },
    {
      "id": "392547",
      "postDate": "09/24/2018 00:45:20",
      "content": "<p>In the vast majority of cases (~98%), <code>totals.transactionRevenue</code> is null, which the contest administrators have clarified means no sale (i.e., 0). Is this what you could be running in to?</p>",
      "rawMarkdown": "In the vast majority of cases (~98%), `totals.transactionRevenue` is null, which the contest administrators have clarified means no sale (i.e., 0). Is this what you could be running in to?",
      "votes": null
    },
    {
      "id": "392584",
      "postDate": "09/24/2018 03:04:05",
      "content": "<p>That solves my issue. Thanks!</p>",
      "rawMarkdown": "That solves my issue. Thanks!",
      "votes": null
    },
    {
      "id": "415906",
      "postDate": "11/05/2018 22:15:08",
      "content": "<p>How do i split these values?</p>",
      "rawMarkdown": "How do i split these values?",
      "votes": null
    },
    {
      "id": "417210",
      "postDate": "11/08/2018 00:10:51",
      "content": "<p>The values here are stored as a string, but look like a dictionary.  You can do something like this to convert it from a string to a dictionary, and then select the key value pairs you want. </p>\n\n<p>train.totals.apply(lambda x: ast.literal_eval(x))</p>",
      "rawMarkdown": "The values here are stored as a string, but look like a dictionary.  You can do something like this to convert it from a string to a dictionary, and then select the key value pairs you want. \n\ntrain.totals.apply(lambda x: ast.literal_eval(x))",
      "votes": null
    }
  ],
  "comments": [
    {
      "id": 392546,
      "author_name": "peterhurford",
      "author_url": "",
      "post_date": "09/24/2018 00:39:44",
      "content": "<p><img src=\"https://i.imgur.com/u2N1Y8r.png\" alt=\"enter image description here\"></p>",
      "votes": null,
      "replies": [
        {
          "id": 392547,
          "author_name": "nnnnick",
          "author_url": "",
          "post_date": "09/24/2018 00:45:20",
          "content": "<p>In the vast majority of cases (~98%), <code>totals.transactionRevenue</code> is null, which the contest administrators have clarified means no sale (i.e., 0). Is this what you could be running in to?</p>",
          "votes": null,
          "replies": []
        },
        {
          "id": 392584,
          "author_name": "peterhurford",
          "author_url": "",
          "post_date": "09/24/2018 03:04:05",
          "content": "<p>That solves my issue. Thanks!</p>",
          "votes": null,
          "replies": []
        }
      ]
    },
    {
      "id": 415906,
      "author_name": "anuraghakeem",
      "author_url": "",
      "post_date": "11/05/2018 22:15:08",
      "content": "<p>How do i split these values?</p>",
      "votes": null,
      "replies": [
        {
          "id": 417210,
          "author_name": "zakraicik",
          "author_url": "",
          "post_date": "11/08/2018 00:10:51",
          "content": "<p>The values here are stored as a string, but look like a dictionary.  You can do something like this to convert it from a string to a dictionary, and then select the key value pairs you want. </p>\n\n<p>train.totals.apply(lambda x: ast.literal_eval(x))</p>",
          "votes": null,
          "replies": []
        }
      ]
    }
  ],
  "raw_markdown_by_id": {
    "392545": "Forgive me for being an idiot, but when I look at `train.csv` on https://www.kaggle.com/c/ga-customer-revenue-prediction/data and it says:\n\n&gt; There are multiple columns which contain JSON blobs of varying depth. In one of those JSON columns, totals, the sub-column transactionRevenue contains the revenue information we are trying to predict. This sub-column exists only for the training data.\n\n...but I don't see that subcolumn anywhere! &gt;.&lt;\n\nWhat am I missing?",
    "392546": "![enter image description here][1]\n\n  [1]: https://i.imgur.com/u2N1Y8r.png",
    "392547": "In the vast majority of cases (~98%), `totals.transactionRevenue` is null, which the contest administrators have clarified means no sale (i.e., 0). Is this what you could be running in to?",
    "392584": "That solves my issue. Thanks!",
    "415906": "How do i split these values?",
    "417210": "The values here are stored as a string, but look like a dictionary.  You can do something like this to convert it from a string to a dictionary, and then select the key value pairs you want. \n\ntrain.totals.apply(lambda x: ast.literal_eval(x))"
  },
  "source": "meta"
}