{ "id": 369921, "title": "Scores update to reflect metric patch", "url": "/competitions/rsna-breast-cancer-detection/discussion/369921", "author_name": "Sohier Dane", "post_date": "2022-12-02T00:40:02.483000", "votes": 21, "comment_count": 0, "views": 0, "content": "

I accidentally included an extra term in my initial implementation of the metric, which I really appreciate @level14taken catching here I've patched the deployed copy of the metric and rescored all existing submissions so you may notice a change in your older scores. I don't expect it to be large in most cases; the sample submission score went from 0.01 to 0.02.

\n

I do apologize for this mistake and hope that it won't be too much of an inconvenience thanks to the swift report by @level14taken.

", "messages": [ { "id": 2052138, "postDate": "2022-12-02T00:40:02.483Z", "content": "

I accidentally included an extra term in my initial implementation of the metric, which I really appreciate @level14taken catching here I've patched the deployed copy of the metric and rescored all existing submissions so you may notice a change in your older scores. I don't expect it to be large in most cases; the sample submission score went from 0.01 to 0.02.

\n

I do apologize for this mistake and hope that it won't be too much of an inconvenience thanks to the swift report by @level14taken.

", "rawMarkdown": "I accidentally included an extra term in my initial implementation of the metric, which I really appreciate @level14taken [catching here](https://www.kaggle.com/code/sohier/probabilistic-f-score/comments#2050902. ) I've patched the deployed copy of the metric and rescored all existing submissions so you may notice a change in your older scores. I don't expect it to be large in most cases; the sample submission score went from 0.01 to 0.02.\n\nI do apologize for this mistake and hope that it won't be too much of an inconvenience thanks to the swift report by @level14taken.", "votes": 21 } ], "comments": [], "raw_markdown_by_id": { "2052138": "I accidentally included an extra term in my initial implementation of the metric, which I really appreciate @level14taken [catching here](https://www.kaggle.com/code/sohier/probabilistic-f-score/comments#2050902. ) I've patched the deployed copy of the metric and rescored all existing submissions so you may notice a change in your older scores. I don't expect it to be large in most cases; the sample submission score went from 0.01 to 0.02.\n\nI do apologize for this mistake and hope that it won't be too much of an inconvenience thanks to the swift report by @level14taken." } }