{
  "id": 42808,
  "title": "Test set have store_item combinations which don't appear in Train set ",
  "url": "/competitions/favorita-grocery-sales-forecasting/discussion/42808",
  "author_name": "",
  "post_date": "2017-11-05T12:29:12.195388900Z",
  "votes": 1,
  "comment_count": 3,
  "views": 0,
  "content": "<p><strong>Test</strong> data set ask to forcast : item_nbr = 103501 and store_nbr = 1 \nbut train doesn't have this combination</p>\n\n<p>while <strong>Train</strong> data set only has 1 record where item_nbr = 2124052 and store_nbr = 54,\nthis will make predict in Test more difficult  </p>",
  "messages": [
    {
      "id": "240034",
      "postDate": "11/05/2017 12:29:12",
      "content": "<p><strong>Test</strong> data set ask to forcast : item_nbr = 103501 and store_nbr = 1 \nbut train doesn't have this combination</p>\n\n<p>while <strong>Train</strong> data set only has 1 record where item_nbr = 2124052 and store_nbr = 54,\nthis will make predict in Test more difficult  </p>",
      "rawMarkdown": "**Test** data set ask to forcast : item_nbr = 103501 and store_nbr = 1 \nbut train doesn't have this combination\n\nwhile **Train** data set only has 1 record where item_nbr = 2124052 and store_nbr = 54,\nthis will make predict in Test more difficult",
      "votes": null
    },
    {
      "id": "240194",
      "postDate": "11/05/2017 22:01:59",
      "content": "<p>Please see the data description; it's done on purpose.</p>\n\n<p><em>NOTE: The test data has a small number of items that are not contained in the training data. Part of the exercise will be to predict a new item sales based on similar products..</em></p>",
      "rawMarkdown": "Please see the data description; it's done on purpose.\n\n*NOTE: The test data has a small number of items that are not contained in the training data. Part of the exercise will be to predict a new item sales based on similar products..*",
      "votes": null
    },
    {
      "id": "240227",
      "postDate": "11/06/2017 01:19:33",
      "content": "<p>right, I should read carefully, while these cases are really we should take care about </p>",
      "rawMarkdown": "right, I should read carefully, while these cases are really we should take care about",
      "votes": null
    },
    {
      "id": "241652",
      "postDate": "11/09/2017 14:03:24",
      "content": "<p>And there are more than 40k item/store combo in that case. See this kernel: <a href=\"https://www.kaggle.com/ericdfr/new-items-in-stores/\">https://www.kaggle.com/ericdfr/new-items-in-stores/</a>. <br>\nThis is huge and not common while the median for these new item/store combo is 39 per day!!!</p>",
      "rawMarkdown": "And there are more than 40k item/store combo in that case. See this kernel: [https://www.kaggle.com/ericdfr/new-items-in-stores/][1].  \nThis is huge and not common while the median for these new item/store combo is 39 per day!!!\n\n\n  [1]: https://www.kaggle.com/ericdfr/new-items-in-stores/",
      "votes": null
    }
  ],
  "comments": [
    {
      "id": 240194,
      "author_name": "flygare",
      "author_url": "",
      "post_date": "11/05/2017 22:01:59",
      "content": "<p>Please see the data description; it's done on purpose.</p>\n\n<p><em>NOTE: The test data has a small number of items that are not contained in the training data. Part of the exercise will be to predict a new item sales based on similar products..</em></p>",
      "votes": null,
      "replies": [
        {
          "id": 240227,
          "author_name": "yanghuiming",
          "author_url": "",
          "post_date": "11/06/2017 01:19:33",
          "content": "<p>right, I should read carefully, while these cases are really we should take care about </p>",
          "votes": null,
          "replies": []
        }
      ]
    },
    {
      "id": 241652,
      "author_name": "ericdfr",
      "author_url": "",
      "post_date": "11/09/2017 14:03:24",
      "content": "<p>And there are more than 40k item/store combo in that case. See this kernel: <a href=\"https://www.kaggle.com/ericdfr/new-items-in-stores/\">https://www.kaggle.com/ericdfr/new-items-in-stores/</a>. <br>\nThis is huge and not common while the median for these new item/store combo is 39 per day!!!</p>",
      "votes": null,
      "replies": []
    }
  ],
  "raw_markdown_by_id": {
    "240034": "**Test** data set ask to forcast : item_nbr = 103501 and store_nbr = 1 \nbut train doesn't have this combination\n\nwhile **Train** data set only has 1 record where item_nbr = 2124052 and store_nbr = 54,\nthis will make predict in Test more difficult",
    "240194": "Please see the data description; it's done on purpose.\n\n*NOTE: The test data has a small number of items that are not contained in the training data. Part of the exercise will be to predict a new item sales based on similar products..*",
    "240227": "right, I should read carefully, while these cases are really we should take care about",
    "241652": "And there are more than 40k item/store combo in that case. See this kernel: [https://www.kaggle.com/ericdfr/new-items-in-stores/][1].  \nThis is huge and not common while the median for these new item/store combo is 39 per day!!!\n\n\n  [1]: https://www.kaggle.com/ericdfr/new-items-in-stores/"
  },
  "source": "meta"
}