![\"\"](\"https://www.googleapis.com/download/storage/v1/b/kaggle-forum-message-attachments/o/inbox%2F3842406%2Fdd02b2b7504ba3536a417e5b712ceda5%2FScreenshot%202025-04-02%20170515.png?generation=1743606354678939&alt=media\")
is there a shift for the spots in slide # 2 of the training set?
\n
is there a shift for the spots in slide # 2 of the training set?
\n
See this other discussion.
\nIn particular, this image from this notebook. It shows a spot that clearly maps to an empty space, yet according to the data, contains a lot of C1 and C3 cells.
\n![\"\"](\"https://www.googleapis.com/download/storage/v1/b/kaggle-forum-message-attachments/o/inbox%2F650%2Fd52280577ad0c8424680e27623c74c29%2FScreenshot%202025-04-02%20162640.png?generation=1743607672151998&alt=media\")
In this other notebook, at the end, I plotted the distribution of spots over all the slices.
\nIt seems slice S_1 and S_2 are shifted. The others may be ok.
I've tried shifting slice S_1 and S_2 manually. I think this could be a good realignment:
\n![\"\"](\"https://www.googleapis.com/download/storage/v1/b/kaggle-forum-message-attachments/o/inbox%2F650%2F89d81332aa44a894f1bd2907f9cd4124%2FScreenshot%202025-04-03%20063820.png?generation=1743658764957888&alt=media\")
\n![\"\"](\"https://www.googleapis.com/download/storage/v1/b/kaggle-forum-message-attachments/o/inbox%2F650%2Fe2d3a0ff90247e5a8ec00153a8f9ec8b%2FScreenshot%202025-04-03%20063743.png?generation=1743658750847772&alt=media\")
does this have all the points correct i mean some points are still outside the image do they show 0 cell concentration?
", "rawMarkdown": "does this have all the points correct i mean some points are still outside the image do they show 0 cell concentration?", "votes": null }, { "id": "3170247", "postDate": "04/04/2025 13:46:06", "content": "It seems to be ok to me, but if you want to have a look and copy my notebook and check it, I'd appreciate it.
", "rawMarkdown": "It seems to be ok to me, but if you want to have a look and copy my notebook and check it, I'd appreciate it.", "votes": null }, { "id": "3170801", "postDate": "04/05/2025 03:03:29", "content": "Real thanks man your notebook is the only public one which took the image into consideration i got tired of thoose which just took x and y you are the only one who is contributing to the public in this competetion i hope you get a good rank
", "rawMarkdown": "Real thanks man your notebook is the only public one which took the image into consideration i got tired of thoose which just took x and y you are the only one who is contributing to the public in this competetion i hope you get a good rank", "votes": null }, { "id": "3171121", "postDate": "04/05/2025 10:16:43", "content": "Thanks, I appreciate that :-)
", "rawMarkdown": "Thanks, I appreciate that :-)", "votes": null } ], "comments": [ { "id": 3168691, "author_name": "dalloliogm", "author_url": "", "post_date": "04/02/2025 17:07:51", "content": "See this other discussion.
\nIn particular, this image from this notebook. It shows a spot that clearly maps to an empty space, yet according to the data, contains a lot of C1 and C3 cells.
\n
In this other notebook, at the end, I plotted the distribution of spots over all the slices.
\nIt seems slice S_1 and S_2 are shifted. The others may be ok.
I've tried shifting slice S_1 and S_2 manually. I think this could be a good realignment:
\n
\n
does this have all the points correct i mean some points are still outside the image do they show 0 cell concentration?
", "votes": null, "replies": [ { "id": 3170247, "author_name": "dalloliogm", "author_url": "", "post_date": "04/04/2025 13:46:06", "content": "It seems to be ok to me, but if you want to have a look and copy my notebook and check it, I'd appreciate it.
", "votes": null, "replies": [ { "id": 3170801, "author_name": "saikotisana", "author_url": "", "post_date": "04/05/2025 03:03:29", "content": "Real thanks man your notebook is the only public one which took the image into consideration i got tired of thoose which just took x and y you are the only one who is contributing to the public in this competetion i hope you get a good rank
", "votes": null, "replies": [ { "id": 3171121, "author_name": "dalloliogm", "author_url": "", "post_date": "04/05/2025 10:16:43", "content": "Thanks, I appreciate that :-)
", "votes": null, "replies": [] } ] } ] } ] } ] } ], "raw_markdown_by_id": { "3168572": "is there a shift for the spots in slide # 2 of the training set?\n\n", "3168691": "See [this ](https://www.kaggle.com/competitions/el-hackathon-2025/discussion/569132)other discussion.\n\nIn particular, this image from this [notebook](https://www.kaggle.com/code/dalloliogm/basic-prediction-using-cnn-and-image-patches). It shows a spot that clearly maps to an empty space, yet according to the data, contains a lot of C1 and C3 cells.\n\n", "3168695": "In this [other ](https://www.kaggle.com/code/dalloliogm/eda-plotting-cell-distribution-by-slice#PLotting-Total-Abundance)notebook, at the end, I plotted the distribution of spots over all the slices.\nIt seems slice S_1 and S_2 are shifted. The others may be ok.", "3169095": "I've tried shifting slice S_1 and S_2 manually. I think this could be a good realignment:\n\n", "3170217": "does this have all the points correct i mean some points are still outside the image do they show 0 cell concentration?", "3170247": "It seems to be ok to me, but if you want to have a look and copy my notebook and check it, I'd appreciate it.", "3170801": "Real thanks man your notebook is the only public one which took the image into consideration i got tired of thoose which just took x and y you are the only one who is contributing to the public in this competetion i hope you get a good rank", "3171121": "Thanks, I appreciate that :-)" }, "source": "meta" }