{ "id": 417655, "title": "anyone successfuly utilize the tile_meta.csv in test set?", "url": "/competitions/hubmap-hacking-the-human-vasculature/discussion/417655", "author_name": "Chenglu", "post_date": "2023-06-16T15:30:13.962000", "votes": 1, "comment_count": 4, "views": 0, "content": "

\n

\n

\n

\n

I didn't read the comment carefully, the tile_meta.csv contains both training and testing annotations, i thought it has only testing set.

", "messages": [ { "id": 2305313, "postDate": "2023-06-16T15:30:13.963Z", "content": "

\n

\n

\n

\n

I didn't read the comment carefully, the tile_meta.csv contains both training and testing annotations, i thought it has only testing set.

", "rawMarkdown": "~~I have had some testing but it keeps failing me during the evaluation.~~\n\n~~My code runs successfully when inferencing with the training set, cause the `tile_meta.csv` is for training.~~\n\n~~Could the host double-check if the `tile_meta.csv` is the right version when doing test set evaluation?~~\n\n~~@ryanholbrook ~~\n\nI didn't read the comment carefully, the `tile_meta.csv` contains both training and testing annotations, i thought it has only testing set. ", "votes": 1 }, { "id": 2305370, "postDate": "2023-06-16T16:25:45.043Z", "content": "

I can confirm that all of the test ids are present in tile_meta.csv during evaluation. I don't see that there's anything wrong with the data either.

", "rawMarkdown": "I can confirm that all of the test ids are present in `tile_meta.csv` during evaluation. I don't see that there's anything wrong with the data either.", "replies": [ { "id": 2305376, "postDate": "2023-06-16T16:38:35.493Z", "content": "

sorry my bad, I thought the tile_meta.csv only contains test set, while It contains both test and training set as you commented in the other thread.

", "rawMarkdown": "sorry my bad, I thought the `tile_meta.csv` only contains test set, while It contains both test and training set as you commented in the other thread. ", "votes": 1 } ] }, { "id": 2305346, "postDate": "2023-06-16T16:05:41.323Z", "content": "

Thanks for the heads up. I'll take a look.

", "rawMarkdown": "Thanks for the heads up. I'll take a look." }, { "id": 2358033, "postDate": "2023-07-25T09:10:23.157Z", "content": "

Here's work from Code and Disscussion Sections
\nhttps://www.kaggle.com/competitions/hubmap-hacking-the-human-vasculature/discussion/422250
\nhttps://www.kaggle.com/code/itsuki9180/investigate-tiles-on-wsi-1-and-2
\nI think they are helpful and data augmentation by cropping small images from the whole image is useful for me.

", "rawMarkdown": "Here's work from Code and Disscussion Sections\nhttps://www.kaggle.com/competitions/hubmap-hacking-the-human-vasculature/discussion/422250\nhttps://www.kaggle.com/code/itsuki9180/investigate-tiles-on-wsi-1-and-2 \nI think they are helpful and data augmentation by cropping small images from the whole image is useful for me. ", "isDeleted": true } ], "comments": [ { "id": 2305370, "author_name": "Ryan Holbrook", "author_url": "", "post_date": "2023-06-16T16:25:45.043000", "content": "

I can confirm that all of the test ids are present in tile_meta.csv during evaluation. I don't see that there's anything wrong with the data either.

", "votes": 0, "replies": [ { "id": 2305376, "author_name": "Chenglu", "author_url": "", "post_date": "2023-06-16T16:38:35.493000", "content": "

sorry my bad, I thought the tile_meta.csv only contains test set, while It contains both test and training set as you commented in the other thread.

", "votes": 1, "replies": [] } ] }, { "id": 2305346, "author_name": "Ryan Holbrook", "author_url": "", "post_date": "2023-06-16T16:05:41.323000", "content": "

Thanks for the heads up. I'll take a look.

", "votes": 0, "replies": [] }, { "id": 2358033, "author_name": "", "author_url": "", "post_date": "2023-07-25T09:10:23.157000", "content": "

Here's work from Code and Disscussion Sections
\nhttps://www.kaggle.com/competitions/hubmap-hacking-the-human-vasculature/discussion/422250
\nhttps://www.kaggle.com/code/itsuki9180/investigate-tiles-on-wsi-1-and-2
\nI think they are helpful and data augmentation by cropping small images from the whole image is useful for me.

", "votes": 0, "replies": [] } ], "raw_markdown_by_id": { "2305313": "~~I have had some testing but it keeps failing me during the evaluation.~~\n\n~~My code runs successfully when inferencing with the training set, cause the `tile_meta.csv` is for training.~~\n\n~~Could the host double-check if the `tile_meta.csv` is the right version when doing test set evaluation?~~\n\n~~@ryanholbrook ~~\n\nI didn't read the comment carefully, the `tile_meta.csv` contains both training and testing annotations, i thought it has only testing set. ", "2305370": "I can confirm that all of the test ids are present in `tile_meta.csv` during evaluation. I don't see that there's anything wrong with the data either.", "2305346": "Thanks for the heads up. I'll take a look.", "2358033": "Here's work from Code and Disscussion Sections\nhttps://www.kaggle.com/competitions/hubmap-hacking-the-human-vasculature/discussion/422250\nhttps://www.kaggle.com/code/itsuki9180/investigate-tiles-on-wsi-1-and-2 \nI think they are helpful and data augmentation by cropping small images from the whole image is useful for me. " } }