{
  "id": 71178,
  "title": "how to get 0.0 score in public LB use test_v2.csv",
  "url": "/competitions/ga-customer-revenue-prediction/discussion/71178",
  "author_name": "",
  "post_date": "2018-11-11T04:24:38.055475700Z",
  "votes": null,
  "comment_count": 6,
  "views": 0,
  "content": "<p>my code as follow</p>\n\n<pre>import numpy as np\nimport pandas as pd\nimport json\nfrom pandas.io.json import json_normalize\nPATH=\"../input/ga-customer-revenue-prediction/\"\ntest_df=pd.read_csv(PATH+\"test_v2.csv\", converters={'totals': json.loads}, dtype={'fullVisitorId': 'str'},usecols=['fullVisitorId','totals'])\ntest_df.head()\n</pre>\n\n<pre>json_col_df = json_normalize(test_df['totals'])\njson_col_df.columns = [f\"totals.{sub_col}\" for sub_col in json_col_df.columns]\ntest_df = test_df.drop('totals', axis=1).merge(json_col_df, right_index=True, left_index=True)\ntest_df.head()\n</pre>\n\n<pre>test_df.shape\n</pre>\n\n<pre>testsub=test_df[['fullVisitorId','totals.transactionRevenue']].fillna(0).groupby('fullVisitorId').sum()\ntestsub.head()\n</pre>\n\n<pre>testsub['fullVisitorId']=testsub.index\ntestsub['PredictedLogRevenue']=testsub['totals.transactionRevenue'].apply(np.log1p)\ntestsub[['fullVisitorId','PredictedLogRevenue']].head()\n</pre>\n\n<pre>testsub[['fullVisitorId','PredictedLogRevenue']].fillna(0).to_csv(\"testsub.csv\",index=False)\n</pre>\n\n<p>I commit it ,but score is always 1.7768  .</p><p>\nDoes something wrong in the code, has idea?</p>",
  "messages": [
    {
      "id": "419004",
      "postDate": "11/11/2018 04:24:38",
      "content": "<p>my code as follow</p>\n\n<pre>import numpy as np\nimport pandas as pd\nimport json\nfrom pandas.io.json import json_normalize\nPATH=\"../input/ga-customer-revenue-prediction/\"\ntest_df=pd.read_csv(PATH+\"test_v2.csv\", converters={'totals': json.loads}, dtype={'fullVisitorId': 'str'},usecols=['fullVisitorId','totals'])\ntest_df.head()\n</pre>\n\n<pre>json_col_df = json_normalize(test_df['totals'])\njson_col_df.columns = [f\"totals.{sub_col}\" for sub_col in json_col_df.columns]\ntest_df = test_df.drop('totals', axis=1).merge(json_col_df, right_index=True, left_index=True)\ntest_df.head()\n</pre>\n\n<pre>test_df.shape\n</pre>\n\n<pre>testsub=test_df[['fullVisitorId','totals.transactionRevenue']].fillna(0).groupby('fullVisitorId').sum()\ntestsub.head()\n</pre>\n\n<pre>testsub['fullVisitorId']=testsub.index\ntestsub['PredictedLogRevenue']=testsub['totals.transactionRevenue'].apply(np.log1p)\ntestsub[['fullVisitorId','PredictedLogRevenue']].head()\n</pre>\n\n<pre>testsub[['fullVisitorId','PredictedLogRevenue']].fillna(0).to_csv(\"testsub.csv\",index=False)\n</pre>\n\n<p>I commit it ,but score is always 1.7768  .</p><p>\nDoes something wrong in the code, has idea?</p>",
      "rawMarkdown": "my code as follow\n<pre>import numpy as np\nimport pandas as pd\nimport json\nfrom pandas.io.json import json_normalize\nPATH=\"../input/ga-customer-revenue-prediction/\"\ntest_df=pd.read_csv(PATH+\"test_v2.csv\", converters={'totals': json.loads}, dtype={'fullVisitorId': 'str'},usecols=['fullVisitorId','totals'])\ntest_df.head()\n</pre>\n<pre>json_col_df = json_normalize(test_df['totals'])\njson_col_df.columns = [f\"totals.{sub_col}\" for sub_col in json_col_df.columns]\ntest_df = test_df.drop('totals', axis=1).merge(json_col_df, right_index=True, left_index=True)\ntest_df.head()\n</pre>\n<pre>test_df.shape\n</pre>\n<pre>testsub=test_df[['fullVisitorId','totals.transactionRevenue']].fillna(0).groupby('fullVisitorId').sum()\ntestsub.head()\n</pre>\n<pre>testsub['fullVisitorId']=testsub.index\ntestsub['PredictedLogRevenue']=testsub['totals.transactionRevenue'].apply(np.log1p)\ntestsub[['fullVisitorId','PredictedLogRevenue']].head()\n</pre>\n<pre>testsub[['fullVisitorId','PredictedLogRevenue']].fillna(0).to_csv(\"testsub.csv\",index=False)\n</pre>\nI commit it ,but score is always 1.7768  .<p>\nDoes something wrong in the code, has idea?\n</p>",
      "votes": null
    },
    {
      "id": "419006",
      "postDate": "11/11/2018 04:33:44",
      "content": "<p>what happened？ I can't submit！</p>",
      "rawMarkdown": "what happened？ I can't submit！",
      "votes": null
    },
    {
      "id": "419008",
      "postDate": "11/11/2018 04:36:05",
      "content": "<p>rule is change.</p><p>\ndata is change.</p><p>\nsubmmit is change.</p>",
      "rawMarkdown": "rule is change.<p>\ndata is change.</p><p>\nsubmmit is change.</p>",
      "votes": null
    },
    {
      "id": "419009",
      "postDate": "11/11/2018 04:38:41",
      "content": "<p>How to submit?</p>",
      "rawMarkdown": "How to submit?",
      "votes": null
    },
    {
      "id": "419013",
      "postDate": "11/11/2018 04:46:35",
      "content": "<p>Can you submit your result?</p>",
      "rawMarkdown": "Can you submit your result?",
      "votes": null
    },
    {
      "id": "419120",
      "postDate": "11/11/2018 10:20:05",
      "content": "<p>I can submit.But score is 1.7768 .</p>",
      "rawMarkdown": "I can submit.But score is 1.7768 .",
      "votes": null
    },
    {
      "id": "419127",
      "postDate": "11/11/2018 10:58:09",
      "content": "<p>see <a href=\"https://www.kaggle.com/paulorzp/1line-perfect-score\">https://www.kaggle.com/paulorzp/1line-perfect-score</a></p><p> \nmaybe json.loads and json_normalize has problem.</p><p></p>",
      "rawMarkdown": "see https://www.kaggle.com/paulorzp/1line-perfect-score<p> \nmaybe json.loads and json_normalize has problem.</p><p></p>",
      "votes": null
    }
  ],
  "comments": [
    {
      "id": 419006,
      "author_name": "dinglubjut25079",
      "author_url": "",
      "post_date": "11/11/2018 04:33:44",
      "content": "<p>what happened？ I can't submit！</p>",
      "votes": null,
      "replies": []
    },
    {
      "id": 419008,
      "author_name": "zjp2origin",
      "author_url": "",
      "post_date": "11/11/2018 04:36:05",
      "content": "<p>rule is change.</p><p>\ndata is change.</p><p>\nsubmmit is change.</p>",
      "votes": null,
      "replies": []
    },
    {
      "id": 419009,
      "author_name": "dinglubjut25079",
      "author_url": "",
      "post_date": "11/11/2018 04:38:41",
      "content": "<p>How to submit?</p>",
      "votes": null,
      "replies": []
    },
    {
      "id": 419013,
      "author_name": "dinglubjut25079",
      "author_url": "",
      "post_date": "11/11/2018 04:46:35",
      "content": "<p>Can you submit your result?</p>",
      "votes": null,
      "replies": [
        {
          "id": 419120,
          "author_name": "zjp2origin",
          "author_url": "",
          "post_date": "11/11/2018 10:20:05",
          "content": "<p>I can submit.But score is 1.7768 .</p>",
          "votes": null,
          "replies": []
        }
      ]
    },
    {
      "id": 419127,
      "author_name": "zjp2origin",
      "author_url": "",
      "post_date": "11/11/2018 10:58:09",
      "content": "<p>see <a href=\"https://www.kaggle.com/paulorzp/1line-perfect-score\">https://www.kaggle.com/paulorzp/1line-perfect-score</a></p><p> \nmaybe json.loads and json_normalize has problem.</p><p></p>",
      "votes": null,
      "replies": []
    }
  ],
  "raw_markdown_by_id": {
    "419004": "my code as follow\n<pre>import numpy as np\nimport pandas as pd\nimport json\nfrom pandas.io.json import json_normalize\nPATH=\"../input/ga-customer-revenue-prediction/\"\ntest_df=pd.read_csv(PATH+\"test_v2.csv\", converters={'totals': json.loads}, dtype={'fullVisitorId': 'str'},usecols=['fullVisitorId','totals'])\ntest_df.head()\n</pre>\n<pre>json_col_df = json_normalize(test_df['totals'])\njson_col_df.columns = [f\"totals.{sub_col}\" for sub_col in json_col_df.columns]\ntest_df = test_df.drop('totals', axis=1).merge(json_col_df, right_index=True, left_index=True)\ntest_df.head()\n</pre>\n<pre>test_df.shape\n</pre>\n<pre>testsub=test_df[['fullVisitorId','totals.transactionRevenue']].fillna(0).groupby('fullVisitorId').sum()\ntestsub.head()\n</pre>\n<pre>testsub['fullVisitorId']=testsub.index\ntestsub['PredictedLogRevenue']=testsub['totals.transactionRevenue'].apply(np.log1p)\ntestsub[['fullVisitorId','PredictedLogRevenue']].head()\n</pre>\n<pre>testsub[['fullVisitorId','PredictedLogRevenue']].fillna(0).to_csv(\"testsub.csv\",index=False)\n</pre>\nI commit it ,but score is always 1.7768  .<p>\nDoes something wrong in the code, has idea?\n</p>",
    "419006": "what happened？ I can't submit！",
    "419008": "rule is change.<p>\ndata is change.</p><p>\nsubmmit is change.</p>",
    "419009": "How to submit?",
    "419013": "Can you submit your result?",
    "419120": "I can submit.But score is 1.7768 .",
    "419127": "see https://www.kaggle.com/paulorzp/1line-perfect-score<p> \nmaybe json.loads and json_normalize has problem.</p><p></p>"
  },
  "source": "meta"
}