{ "id": 335507, "title": "Need Help.... Submission Score Error.... ?", "url": "/competitions/hubmap-organ-segmentation/discussion/335507", "author_name": "", "post_date": "2022-07-06T13:42:35.640670700Z", "votes": 7, "comment_count": 10, "views": 0, "content": "

Here is my submission notebook run for a long time and returens submission score error.
\nhttps://www.kaggle.com/code/muki2003/submission-hubmap-hpa-u-net-cnn

\n

what I have done wrong ?

", "messages": [ { "id": "1845649", "postDate": "07/06/2022 13:42:35", "content": "

Here is my submission notebook run for a long time and returens submission score error.
\nhttps://www.kaggle.com/code/muki2003/submission-hubmap-hpa-u-net-cnn

\n

what I have done wrong ?

", "rawMarkdown": "Here is my submission notebook run for a long time and returens submission score error.\nhttps://www.kaggle.com/code/muki2003/submission-hubmap-hpa-u-net-cnn\n\nwhat I have done wrong ?", "votes": null }, { "id": "1845933", "postDate": "07/06/2022 17:30:37", "content": "

Hi,

\n

I think the problem could be that the memory is not big enough to load the raw tiff file.
\nIn the data description, they said:

\n

\"Expect roughly 550 images in the hidden test set. All HPA images are 3000 x 3000 pixels with a tissue area within the image around 2500 x 2500 pixels. The HuBMAP images range in size from 4500x4500 down to 160x160 pixels.\"

", "rawMarkdown": "Hi, \n\nI think the problem could be that the memory is not big enough to load the raw tiff file.\nIn the data description, they said: \n\n\"Expect roughly 550 images in the hidden test set. All HPA images are 3000 x 3000 pixels with a tissue area within the image around 2500 x 2500 pixels. The HuBMAP images range in size from 4500x4500 down to 160x160 pixels.\"", "votes": null }, { "id": "1845965", "postDate": "07/06/2022 18:09:41", "content": "

Yeah @zhijianli but I am not loaded all image

\n
for i, filename in tqdm(enumerate(glob.glob('../input/hubmap-organ-segmentation/test_images/*.tiff')), \n                        total = len(list(glob.glob('../input/hubmap-organ-segmentation/test_images/*.tiff')))):\n\n    test_image = tiff.imread(filename) / 255.\n    test_image = resize(test_image, (256, 256, 3))\n\n    predicted = model.predict(test_image.reshape(-1,256,256,3))  # Predicting the Imagee\n    predicted = predicted.reshape(256,256)\n\n    final_preds_mask.append(predicted.round()) # Mask\n    final_pred_image.append(test_image) # Image\n
\n

Loading image one by one and resizing to 256x256.

\n

Also, the Error is Submission Score Error

", "rawMarkdown": "Yeah @zhijianli but **I am not loaded all image**\n\n```\nfor i, filename in tqdm(enumerate(glob.glob('../input/hubmap-organ-segmentation/test_images/*.tiff')), \n total = len(list(glob.glob('../input/hubmap-organ-segmentation/test_images/*.tiff')))):\n\n test_image = tiff.imread(filename) / 255.\n test_image = resize(test_image, (256, 256, 3))\n \n predicted = model.predict(test_image.reshape(-1,256,256,3)) # Predicting the Imagee\n predicted = predicted.reshape(256,256)\n \n final_preds_mask.append(predicted.round()) # Mask\n final_pred_image.append(test_image) # Image\n```\n\n**Loading image one by one and resizing to 256x256.**\n\nAlso, the Error is Submission Score Error", "votes": null }, { "id": "1846785", "postDate": "07/07/2022 11:02:05", "content": "

Is the size of mask right? I think the size of mask should be the original shape in test.csv

", "rawMarkdown": "Is the size of mask right? I think the size of mask should be the original shape in test.csv", "votes": null }, { "id": "1846854", "postDate": "07/07/2022 12:02:00", "content": "

@muki2003 Sure I will check it up.

", "rawMarkdown": "muki2003 Sure I will check it up.", "votes": null }, { "id": "1847966", "postDate": "07/08/2022 09:29:40", "content": "

I think this is a simple error on saving the submission. Set the index=False for the save function.

\n

df_ssub.to_csv('submission.csv', index=False)

\n

I encountered such error once before, hope this helps.

", "rawMarkdown": "I think this is a simple error on saving the submission. Set the index=False for the save function. \n\n`df_ssub.to_csv('submission.csv', index=False)`\n\nI encountered such error once before, hope this helps.", "votes": null }, { "id": "1849194", "postDate": "07/09/2022 09:43:14", "content": "

There may be a problem with the function mask2rle? Try rle_encode_less_memory from here
\nhttps://www.kaggle.com/code/realneuralnetwork/fork-of-inference-fastai-baseline

", "rawMarkdown": "There may be a problem with the function mask2rle? Try **rle_encode_less_memory** from here\nhttps://www.kaggle.com/code/realneuralnetwork/fork-of-inference-fastai-baseline", "votes": null }, { "id": "1888182", "postDate": "08/07/2022 12:07:47", "content": "

Hello @muki2003, I am facing the same problem. Have you solved it?

", "rawMarkdown": "Hello @muki2003, I am facing the same problem. Have you solved it?", "votes": null }, { "id": "1888374", "postDate": "08/07/2022 14:14:40", "content": "

Yeah @vominhcong05

\n

Check Out my Submission Notebook
\nhttps://www.kaggle.com/code/muki2003/submission-hubmap-hpa-u-net-cnn

\n

Still you have any issue tag me !

", "rawMarkdown": "Yeah @vominhcong05 \n\nCheck Out my Submission Notebook\nhttps://www.kaggle.com/code/muki2003/submission-hubmap-hpa-u-net-cnn\n\nStill you have any issue tag me !", "votes": null }, { "id": "1890064", "postDate": "08/08/2022 14:21:29", "content": "

Hi @muki2003 , thank you very much. I finally submitted my own submission successfully.
\nI just loaded sample_submission.csv file and then replace rle values with my predicted rle.
\nI still have no idea why it works. It seems that rle dtype should be \"something\" exactly same to \"rle\" dtype from sample_submission.csv.

", "rawMarkdown": "Hi @muki2003 , thank you very much. I finally submitted my own submission successfully. \nI just loaded `sample_submission.csv` file and then replace `rle` values with my predicted rle. \nI still have no idea why it works. It seems that `rle` dtype should be \"something\" exactly same to \"rle\" dtype from `sample_submission.csv`.", "votes": null }, { "id": "1890146", "postDate": "08/08/2022 15:13:08", "content": "

@vominhcong05 Keep going !

", "rawMarkdown": "vominhcong05 Keep going !", "votes": null } ], "comments": [ { "id": 1845933, "author_name": "zhijianli", "author_url": "", "post_date": "07/06/2022 17:30:37", "content": "

Hi,

\n

I think the problem could be that the memory is not big enough to load the raw tiff file.
\nIn the data description, they said:

\n

\"Expect roughly 550 images in the hidden test set. All HPA images are 3000 x 3000 pixels with a tissue area within the image around 2500 x 2500 pixels. The HuBMAP images range in size from 4500x4500 down to 160x160 pixels.\"

", "votes": null, "replies": [ { "id": 1845965, "author_name": "muki2003", "author_url": "", "post_date": "07/06/2022 18:09:41", "content": "

Yeah @zhijianli but I am not loaded all image

\n
for i, filename in tqdm(enumerate(glob.glob('../input/hubmap-organ-segmentation/test_images/*.tiff')), \n                        total = len(list(glob.glob('../input/hubmap-organ-segmentation/test_images/*.tiff')))):\n\n    test_image = tiff.imread(filename) / 255.\n    test_image = resize(test_image, (256, 256, 3))\n\n    predicted = model.predict(test_image.reshape(-1,256,256,3))  # Predicting the Imagee\n    predicted = predicted.reshape(256,256)\n\n    final_preds_mask.append(predicted.round()) # Mask\n    final_pred_image.append(test_image) # Image\n
\n

Loading image one by one and resizing to 256x256.

\n

Also, the Error is Submission Score Error

", "votes": null, "replies": [] } ] }, { "id": 1846785, "author_name": "jiageng", "author_url": "", "post_date": "07/07/2022 11:02:05", "content": "

Is the size of mask right? I think the size of mask should be the original shape in test.csv

", "votes": null, "replies": [ { "id": 1846854, "author_name": "muki2003", "author_url": "", "post_date": "07/07/2022 12:02:00", "content": "

@muki2003 Sure I will check it up.

", "votes": null, "replies": [] } ] }, { "id": 1847966, "author_name": "seraphwedd18", "author_url": "", "post_date": "07/08/2022 09:29:40", "content": "

I think this is a simple error on saving the submission. Set the index=False for the save function.

\n

df_ssub.to_csv('submission.csv', index=False)

\n

I encountered such error once before, hope this helps.

", "votes": null, "replies": [] }, { "id": 1849194, "author_name": "aleksandrkruchinin", "author_url": "", "post_date": "07/09/2022 09:43:14", "content": "

There may be a problem with the function mask2rle? Try rle_encode_less_memory from here
\nhttps://www.kaggle.com/code/realneuralnetwork/fork-of-inference-fastai-baseline

", "votes": null, "replies": [] }, { "id": 1888182, "author_name": "vominhcong05", "author_url": "", "post_date": "08/07/2022 12:07:47", "content": "

Hello @muki2003, I am facing the same problem. Have you solved it?

", "votes": null, "replies": [ { "id": 1888374, "author_name": "muki2003", "author_url": "", "post_date": "08/07/2022 14:14:40", "content": "

Yeah @vominhcong05

\n

Check Out my Submission Notebook
\nhttps://www.kaggle.com/code/muki2003/submission-hubmap-hpa-u-net-cnn

\n

Still you have any issue tag me !

", "votes": null, "replies": [] }, { "id": 1890064, "author_name": "vominhcong05", "author_url": "", "post_date": "08/08/2022 14:21:29", "content": "

Hi @muki2003 , thank you very much. I finally submitted my own submission successfully.
\nI just loaded sample_submission.csv file and then replace rle values with my predicted rle.
\nI still have no idea why it works. It seems that rle dtype should be \"something\" exactly same to \"rle\" dtype from sample_submission.csv.

", "votes": null, "replies": [] }, { "id": 1890146, "author_name": "muki2003", "author_url": "", "post_date": "08/08/2022 15:13:08", "content": "

@vominhcong05 Keep going !

", "votes": null, "replies": [] } ] } ], "raw_markdown_by_id": { "1845649": "Here is my submission notebook run for a long time and returens submission score error.\nhttps://www.kaggle.com/code/muki2003/submission-hubmap-hpa-u-net-cnn\n\nwhat I have done wrong ?", "1845933": "Hi, \n\nI think the problem could be that the memory is not big enough to load the raw tiff file.\nIn the data description, they said: \n\n\"Expect roughly 550 images in the hidden test set. All HPA images are 3000 x 3000 pixels with a tissue area within the image around 2500 x 2500 pixels. The HuBMAP images range in size from 4500x4500 down to 160x160 pixels.\"", "1845965": "Yeah @zhijianli but **I am not loaded all image**\n\n```\nfor i, filename in tqdm(enumerate(glob.glob('../input/hubmap-organ-segmentation/test_images/*.tiff')), \n total = len(list(glob.glob('../input/hubmap-organ-segmentation/test_images/*.tiff')))):\n\n test_image = tiff.imread(filename) / 255.\n test_image = resize(test_image, (256, 256, 3))\n \n predicted = model.predict(test_image.reshape(-1,256,256,3)) # Predicting the Imagee\n predicted = predicted.reshape(256,256)\n \n final_preds_mask.append(predicted.round()) # Mask\n final_pred_image.append(test_image) # Image\n```\n\n**Loading image one by one and resizing to 256x256.**\n\nAlso, the Error is Submission Score Error", "1846785": "Is the size of mask right? I think the size of mask should be the original shape in test.csv", "1846854": "muki2003 Sure I will check it up.", "1847966": "I think this is a simple error on saving the submission. Set the index=False for the save function. \n\n`df_ssub.to_csv('submission.csv', index=False)`\n\nI encountered such error once before, hope this helps.", "1849194": "There may be a problem with the function mask2rle? Try **rle_encode_less_memory** from here\nhttps://www.kaggle.com/code/realneuralnetwork/fork-of-inference-fastai-baseline", "1888182": "Hello @muki2003, I am facing the same problem. Have you solved it?", "1888374": "Yeah @vominhcong05 \n\nCheck Out my Submission Notebook\nhttps://www.kaggle.com/code/muki2003/submission-hubmap-hpa-u-net-cnn\n\nStill you have any issue tag me !", "1890064": "Hi @muki2003 , thank you very much. I finally submitted my own submission successfully. \nI just loaded `sample_submission.csv` file and then replace `rle` values with my predicted rle. \nI still have no idea why it works. It seems that `rle` dtype should be \"something\" exactly same to \"rle\" dtype from `sample_submission.csv`.", "1890146": "vominhcong05 Keep going !" }, "source": "meta" }