{
  "id": 23546,
  "title": "Why all responses in train_numeric.csv are '0'?",
  "url": "/competitions/bosch-production-line-performance/discussion/23546",
  "author_name": "",
  "post_date": "2016-09-10T12:01:49.597Z",
  "votes": null,
  "comment_count": 5,
  "views": 887,
  "content": "<p>According the competition details all defect parts must be &quot;represented by a 'Response' = 1&quot;, but when I dump the Response column from train_numeric.csv  none of the rows has value '1'. How to distinguish failed quality parts from the other?</p>",
  "messages": [
    {
      "id": "135031",
      "postDate": "09/10/2016 12:01:49",
      "content": "<p>According the competition details all defect parts must be &quot;represented by a 'Response' = 1&quot;, but when I dump the Response column from train_numeric.csv  none of the rows has value '1'. How to distinguish failed quality parts from the other?</p>",
      "rawMarkdown": "According the competition details all defect parts must be \"represented by a 'Response' = 1\", but when I dump the Response column from train_numeric.csv  none of the rows has value '1'. How to distinguish failed quality parts from the other?",
      "votes": null
    },
    {
      "id": "135034",
      "postDate": "09/10/2016 12:51:00",
      "content": "<p>You must have made an error</p>",
      "rawMarkdown": "You must have made an error",
      "votes": null
    },
    {
      "id": "135036",
      "postDate": "09/10/2016 13:08:43",
      "content": "<p>Zachary</p>\n\n<p>In case it helps... see this script that shows there are 6879 rows in train_numeric with response =1</p>\n\n<p><a href=\"https://www.kaggle.com/nigelcarpenter/bosch-production-line-performance/response-summary\">https://www.kaggle.com/nigelcarpenter/bosch-production-line-performance/response-summary</a></p>",
      "rawMarkdown": "Zachary\r\n\r\nIn case it helps... see this script that shows there are 6879 rows in train_numeric with response =1\r\n\r\nhttps://www.kaggle.com/nigelcarpenter/bosch-production-line-performance/response-summary",
      "votes": null
    },
    {
      "id": "135413",
      "postDate": "09/13/2016 22:52:57",
      "content": "<p>There are positive Response. I have bin the data by 1000 Ids and count the number of positive Response. Take a look at the picture I have created. </p>",
      "rawMarkdown": "There are positive Response. I have bin the data by 1000 Ids and count the number of positive Response. Take a look at the picture I have created.",
      "votes": null
    },
    {
      "id": "135461",
      "postDate": "09/14/2016 09:32:30",
      "content": "<p>The data is highly skewed [only <strong>0.581 %</strong> of the total set has response 1]  ...so its pretty easy to mis out 0.581 % </p>",
      "rawMarkdown": "The data is highly skewed [only **0.581 %** of the total set has response 1]  ...so its pretty easy to mis out 0.581 %",
      "votes": null
    },
    {
      "id": "137317",
      "postDate": "09/30/2016 10:39:59",
      "content": "<p>Thanks for the answers, actually it was my mistake.</p>",
      "rawMarkdown": "Thanks for the answers, actually it was my mistake.",
      "votes": null
    }
  ],
  "comments": [
    {
      "id": 135034,
      "author_name": "domcastro",
      "author_url": "",
      "post_date": "09/10/2016 12:51:00",
      "content": "<p>You must have made an error</p>",
      "votes": null,
      "replies": []
    },
    {
      "id": 135036,
      "author_name": "nigelcarpenter",
      "author_url": "",
      "post_date": "09/10/2016 13:08:43",
      "content": "<p>Zachary</p>\n\n<p>In case it helps... see this script that shows there are 6879 rows in train_numeric with response =1</p>\n\n<p><a href=\"https://www.kaggle.com/nigelcarpenter/bosch-production-line-performance/response-summary\">https://www.kaggle.com/nigelcarpenter/bosch-production-line-performance/response-summary</a></p>",
      "votes": null,
      "replies": []
    },
    {
      "id": 135413,
      "author_name": "kimktang",
      "author_url": "",
      "post_date": "09/13/2016 22:52:57",
      "content": "<p>There are positive Response. I have bin the data by 1000 Ids and count the number of positive Response. Take a look at the picture I have created. </p>",
      "votes": null,
      "replies": []
    },
    {
      "id": 135461,
      "author_name": "ganeshsivaraman",
      "author_url": "",
      "post_date": "09/14/2016 09:32:30",
      "content": "<p>The data is highly skewed [only <strong>0.581 %</strong> of the total set has response 1]  ...so its pretty easy to mis out 0.581 % </p>",
      "votes": null,
      "replies": []
    },
    {
      "id": 137317,
      "author_name": "zaczac",
      "author_url": "",
      "post_date": "09/30/2016 10:39:59",
      "content": "<p>Thanks for the answers, actually it was my mistake.</p>",
      "votes": null,
      "replies": []
    }
  ],
  "raw_markdown_by_id": {
    "135031": "According the competition details all defect parts must be \"represented by a 'Response' = 1\", but when I dump the Response column from train_numeric.csv  none of the rows has value '1'. How to distinguish failed quality parts from the other?",
    "135034": "You must have made an error",
    "135036": "Zachary\r\n\r\nIn case it helps... see this script that shows there are 6879 rows in train_numeric with response =1\r\n\r\nhttps://www.kaggle.com/nigelcarpenter/bosch-production-line-performance/response-summary",
    "135413": "There are positive Response. I have bin the data by 1000 Ids and count the number of positive Response. Take a look at the picture I have created.",
    "135461": "The data is highly skewed [only **0.581 %** of the total set has response 1]  ...so its pretty easy to mis out 0.581 %",
    "137317": "Thanks for the answers, actually it was my mistake."
  },
  "source": "meta"
}