{
  "id": 477838,
  "title": "although my notebook is running fine im getting scoring error!!",
  "url": "/competitions/home-credit-credit-risk-model-stability/discussion/477838",
  "author_name": "",
  "post_date": "2024-02-18T03:24:15.841001900Z",
  "votes": 1,
  "comment_count": 3,
  "views": 0,
  "content": "<p><code>test['score'] = XGBClassifier().fit(X_train, y_train).predict_proba(X_test)[:, 1]\ntest[['case_id', 'score']].to_csv('submission.csv', index = False)</code> this is my code block</p>",
  "messages": [
    {
      "id": "2656808",
      "postDate": "02/18/2024 03:24:15",
      "content": "<p><code>test['score'] = XGBClassifier().fit(X_train, y_train).predict_proba(X_test)[:, 1]\ntest[['case_id', 'score']].to_csv('submission.csv', index = False)</code> this is my code block</p>",
      "rawMarkdown": "`test['score'] = XGBClassifier().fit(X_train, y_train).predict_proba(X_test)[:, 1]\ntest[['case_id', 'score']].to_csv('submission.csv', index = False)` this is my code block",
      "votes": null
    },
    {
      "id": "2656926",
      "postDate": "02/18/2024 06:42:12",
      "content": "<p>Hello <a href=\"https://www.kaggle.com/townsenddeirective\" target=\"_blank\">@townsenddeirective</a> your notebook is using more RAM than available causing this problem. You can predict in batches and resolve this issue. </p>\n<p>Pseudo-code:-</p>\n<ol>\n<li>Split the test set into very small batches- say 5000 rows per batch</li>\n<li>Predict for 1 batch at a time</li>\n<li>Dump the predictions in the working directory</li>\n<li>Repeat steps 1-3 across all batches to cover the test set</li>\n<li>Import 1 batch at a time and concatenate to create a test set prediction array/ series</li>\n<li>Submit to the LB </li>\n</ol>\n<p>As a tip, I may suggest you to please delete all unassigned and superfluous objects to conserve memory. </p>",
      "rawMarkdown": "Hello @townsenddeirective your notebook is using more RAM than available causing this problem. You can predict in batches and resolve this issue. \n\nPseudo-code:-\n1. Split the test set into very small batches- say 5000 rows per batch\n2. Predict for 1 batch at a time\n3. Dump the predictions in the working directory\n4. Repeat steps 1-3 across all batches to cover the test set\n5. Import 1 batch at a time and concatenate to create a test set prediction array/ series\n6. Submit to the LB \n\nAs a tip, I may suggest you to please delete all unassigned and superfluous objects to conserve memory.",
      "votes": null
    },
    {
      "id": "2657260",
      "postDate": "02/18/2024 12:05:05",
      "content": "<p>I've got a memory error at the beginning then i fixed it and then i was having a scoring error how r they the same if botj have different error name</p>",
      "rawMarkdown": "I've got a memory error at the beginning then i fixed it and then i was having a scoring error how r they the same if botj have different error name",
      "votes": null
    },
    {
      "id": "2695524",
      "postDate": "03/13/2024 17:59:27",
      "content": "<p>did you find a solution? I also have submissions scorring error now and I don’t understand where to look</p>",
      "rawMarkdown": "did you find a solution? I also have submissions scorring error now and I don’t understand where to look",
      "votes": null
    }
  ],
  "comments": [
    {
      "id": 2656926,
      "author_name": "ravi20076",
      "author_url": "",
      "post_date": "02/18/2024 06:42:12",
      "content": "<p>Hello <a href=\"https://www.kaggle.com/townsenddeirective\" target=\"_blank\">@townsenddeirective</a> your notebook is using more RAM than available causing this problem. You can predict in batches and resolve this issue. </p>\n<p>Pseudo-code:-</p>\n<ol>\n<li>Split the test set into very small batches- say 5000 rows per batch</li>\n<li>Predict for 1 batch at a time</li>\n<li>Dump the predictions in the working directory</li>\n<li>Repeat steps 1-3 across all batches to cover the test set</li>\n<li>Import 1 batch at a time and concatenate to create a test set prediction array/ series</li>\n<li>Submit to the LB </li>\n</ol>\n<p>As a tip, I may suggest you to please delete all unassigned and superfluous objects to conserve memory. </p>",
      "votes": null,
      "replies": [
        {
          "id": 2657260,
          "author_name": "townsenddeirective",
          "author_url": "",
          "post_date": "02/18/2024 12:05:05",
          "content": "<p>I've got a memory error at the beginning then i fixed it and then i was having a scoring error how r they the same if botj have different error name</p>",
          "votes": null,
          "replies": [
            {
              "id": 2695524,
              "author_name": "dima1992",
              "author_url": "",
              "post_date": "03/13/2024 17:59:27",
              "content": "<p>did you find a solution? I also have submissions scorring error now and I don’t understand where to look</p>",
              "votes": null,
              "replies": []
            }
          ]
        }
      ]
    }
  ],
  "raw_markdown_by_id": {
    "2656808": "`test['score'] = XGBClassifier().fit(X_train, y_train).predict_proba(X_test)[:, 1]\ntest[['case_id', 'score']].to_csv('submission.csv', index = False)` this is my code block",
    "2656926": "Hello @townsenddeirective your notebook is using more RAM than available causing this problem. You can predict in batches and resolve this issue. \n\nPseudo-code:-\n1. Split the test set into very small batches- say 5000 rows per batch\n2. Predict for 1 batch at a time\n3. Dump the predictions in the working directory\n4. Repeat steps 1-3 across all batches to cover the test set\n5. Import 1 batch at a time and concatenate to create a test set prediction array/ series\n6. Submit to the LB \n\nAs a tip, I may suggest you to please delete all unassigned and superfluous objects to conserve memory.",
    "2657260": "I've got a memory error at the beginning then i fixed it and then i was having a scoring error how r they the same if botj have different error name",
    "2695524": "did you find a solution? I also have submissions scorring error now and I don’t understand where to look"
  },
  "source": "meta"
}