New to machine learning and data science? No question is too basic or too simple. Feel free to start your own thread, or use this thread as a place to post any first-timer clarifying questions for the Kaggle community to help you with!
\nIf you would consider yourself a beginner but don't know where to get started, let other Kagglers help you take your first steps here!
\nNew to Kaggle? Take a look at a few videos Dr. Rachael Tatman has put together to learn a bit more about site etiquette, or Kaggle lingo.
\nRemember: Kaggle is for everyone. Whether you're teaming up or sharing tips in the competition forum, we expect everyone to follow our Kaggle community guidelines.
\nA tip on sharing content - Kaggle is a collaborative community, whereby sharing techniques, starter notebooks, and ideas in the discussion forums are highly encouraged throughout the competition. However, as the competition draws closer to the final deadline it's customary to keep high-scoring notebooks withheld until after the competition has concluded. This maintains the spirit of the competition, while also allowing individuals to submit their own creative work without jeopardy of a higher-scoring notebook being available for an automatic higher rank (through copy/submit). We disable publishing of public notebooks within the final week of the competition, but encourage you to use your best judgment prior to that the deadline.
\nHappy Modeling!
", "messages": [ { "id": 2047812, "postDate": "2022-11-29T04:05:13.717Z", "content": "New to machine learning and data science? No question is too basic or too simple. Feel free to start your own thread, or use this thread as a place to post any first-timer clarifying questions for the Kaggle community to help you with!
\nIf you would consider yourself a beginner but don't know where to get started, let other Kagglers help you take your first steps here!
\nNew to Kaggle? Take a look at a few videos Dr. Rachael Tatman has put together to learn a bit more about site etiquette, or Kaggle lingo.
\nRemember: Kaggle is for everyone. Whether you're teaming up or sharing tips in the competition forum, we expect everyone to follow our Kaggle community guidelines.
\nA tip on sharing content - Kaggle is a collaborative community, whereby sharing techniques, starter notebooks, and ideas in the discussion forums are highly encouraged throughout the competition. However, as the competition draws closer to the final deadline it's customary to keep high-scoring notebooks withheld until after the competition has concluded. This maintains the spirit of the competition, while also allowing individuals to submit their own creative work without jeopardy of a higher-scoring notebook being available for an automatic higher rank (through copy/submit). We disable publishing of public notebooks within the final week of the competition, but encourage you to use your best judgment prior to that the deadline.
\nHappy Modeling!
", "rawMarkdown": "New to machine learning and data science? No question is too basic or too simple. Feel free to start your own thread, or use this thread as a place to post any first-timer clarifying questions for the Kaggle community to help you with!\n\nIf you would consider yourself a beginner but don't know where to get started, let other Kagglers help you take your first steps here!\n\nNew to Kaggle? Take a look at a few videos Dr. Rachael Tatman has put together to learn a bit more about [site etiquette](https://www.youtube.com/watch?v=aIus8si_Et0), or [Kaggle lingo](https://www.youtube.com/watch?v=sEJHyuWKd-s).\n\nRemember: Kaggle is for everyone. Whether you're teaming up or sharing tips in the competition forum, we expect everyone to follow our [Kaggle community guidelines](https://www.kaggle.com/community-guidelines).\n\nA tip on sharing content - Kaggle is a collaborative community, whereby sharing techniques, starter notebooks, and ideas in the discussion forums are highly encouraged throughout the competition. However, as the competition draws closer to the final deadline it's customary to keep high-scoring notebooks withheld until after the competition has concluded. This maintains the spirit of the competition, while also allowing individuals to submit their own creative work without jeopardy of a higher-scoring notebook being available for an automatic higher rank (through copy/submit). We disable publishing of public notebooks within the final week of the competition, but encourage you to use your best judgment prior to that the deadline.\n\n \n\nHappy Modeling!\n", "votes": 15 }, { "id": 2085792, "postDate": "2023-01-04T11:57:52.937Z", "content": "Hi all :)
\nI want to ask a question about downloading data.
\nWhen clicking **Data ** under the ribbon that include :
\nOverview Data Code Discussion Leaderboard Rules Team there is option/ tab that called
\n** Download All** which include 324.72 GB.
\nI do not have so much free space on the laptop I work on.
\nWanted to ask if there is an option to download part of the data?
I'm not sure if related, there is a link to Kaggle API installation and documentation, However I got lost there with unfamiliar terminology, and was sure if it's related to data downloading.
\nKindly your advice or guidance
\nTal
Hi Mohamed,
\nThank you for the detailed answer, much appeiciated!
I followed your steps and when I run the first block of your code in my Jupyter notebook I receive this error:
\nFileNotFoundError: [Errno 2] No such file or directory: 'train.csv'
Do I need to download this file 'train.csv' from somewhere?
\n(because the folder where I run my Jupyter notebook - \"C:\\Users.kaggle\\\" where kaggle.json is)
kindly your advice or guidance
\nTal
thank you very much its much clearer now :)
", "rawMarkdown": "thank you very much its much clearer now :) " }, { "id": 2116413, "postDate": "2023-01-26T13:24:51.863Z", "content": "Hi Mohamed,
\nI have another question with your permission.
\nthe code you wrote in part 3
\n(3) Iterate over the subset and **download every** image separately. My code:
however the first line in the code (which I rum from the jupyter note book, is this correct?)
\ndownloaded_images = [filename for filename in os.listdir() if '.dcm' in filename]
\nseems to assume I already have .dcm files in my folder (which I do not).
Can you please elaborate - how do I get .dcm files in my folder?
\nor is there something else that I should do?
\nThank you in advance
thank you Mohamed 🙏
", "rawMarkdown": "thank you Mohamed 🙏" } ] } ] }, { "id": 2534655, "postDate": "2023-11-22T19:20:05.307Z", "content": "hello everyone I want to work on this competition though its already over but data size if 372 gb i don't have much of space on my laptop is there any option to download part of the data?
", "rawMarkdown": "hello everyone I want to work on this competition though its already over but data size if 372 gb i don't have much of space on my laptop is there any option to download part of the data?\n" }, { "id": 2165020, "postDate": "2023-03-01T23:48:47.210Z", "content": "Any idea on how to incorporate Kaggle newly launched models in while training your model?
\nPreferably using fast ai. Thank you.
Hi,
\nMy question is about the submission and how the evaluation is done.
I'm quit lost in the workflow.
\nThe test data has only 4 files.
\nSo I do not understand where and at what stage all other 8000 patients (from the hidden test set) being added to the model for evaluation?
\nHow do I add all other 8000 patients with their images to the submission.csv file?
(I know I'm missing something not sure what)
\nkindly your help \\ guidance
", "rawMarkdown": "Hi, \nMy question is about the submission and how the evaluation is done.\n\nI'm quit lost in the workflow. \n\nThe test data has only 4 files. \n\nSo I do not understand where and at what stage all other 8000 patients (from the hidden test set) being added to the model for evaluation?\nHow do I add all other 8000 patients with their images to the submission.csv file?\n\n(I know I'm missing something not sure what)\n\nkindly your help \\ guidance \n \n", "replies": [ { "id": 2132226, "postDate": "2023-02-06T17:12:55.417Z", "content": "found the answer here
\nhttps://www.kaggle.com/competitions/rsna-breast-cancer-detection/discussion/382039
Hi there,
\nThis is a newbie kaggle question.
\nI'm running into issues due to not having internet access when making a competition submission.
\nTo this end, I'm looking for some tips on how to load a pretrained resnet50 model in the form of a .pth file.
\nI've copied it over to the appropriate checkpoint folder from /kaggle/input using the code below.
\nif not os.path.exists('/root/.cache/torch/hub/checkpoints/'):
\n os.makedirs('/root/.cache/torch/hub/checkpoints/')
\n!cp '/kaggle/input/resnet50-with-pytorch/resnet50.pth' '/root/.cache/torch/hub/checkpoints/resnet50.pth'
I've also defined a model class, however, I get the following error when I try to instantiate my model class.
\nNameError: name 'resnet50' is not defined
\nAny help or pointers would be greatly appreciated.
\nThank you,
\nBernie
could you post the first lines of your model class and the way you try to load your weights?
", "rawMarkdown": "could you post the first lines of your model class and the way you try to load your weights?\n" } ] }, { "id": 2087605, "postDate": "2023-01-05T17:53:54.597Z", "content": "Hi All. I've used kaggle before, but years ago, and it seems that the set-up has changed. Please could you help asnwer my list of possibly stupid questions:
\nHow do we use open source software available on pip/conda? I've seen some notebooks installing packages from pip, when I try to install something I get : [Errno -3] Temporary failure in name resolution')', is there something else I am supposed to do? - i think I found a way to do this, using something called a wheel file, will update if it works
Can we train a model, save it, and then upload the model and use that to predict on the test data or do we need to submit a notebook that trains the model and then predicts on the test data as the submission?
I saw some mention of a time limit in the discussions, but there's nothing in the rules about this. Is there a time limit for training models or predicting on an image?
Thanks in advance for any help here.
", "rawMarkdown": "Hi All. I've used kaggle before, but years ago, and it seems that the set-up has changed. Please could you help asnwer my list of possibly stupid questions:\n\n1. How do we use open source software available on pip/conda? I've seen some notebooks installing packages from pip, when I try to install something I get : [Errno -3] Temporary failure in name resolution')', is there something else I am supposed to do? - i think I found a way to do this, using something called a wheel file, will update if it works\n\n2. Can we train a model, save it, and then upload the model and use that to predict on the test data or do we need to submit a notebook that trains the model and then predicts on the test data as the submission?\n\n3. I saw some mention of a time limit in the discussions, but there's nothing in the rules about this. Is there a time limit for training models or predicting on an image? \n\nThanks in advance for any help here.", "replies": [ { "id": 2087638, "postDate": "2023-01-05T18:21:36.670Z", "content": "https://www.kaggle.com/competitions/rsna-breast-cancer-detection/overview/code-requirements
", "rawMarkdown": "https://www.kaggle.com/competitions/rsna-breast-cancer-detection/overview/code-requirements", "replies": [ { "id": 2088660, "postDate": "2023-01-06T14:26:26.967Z", "content": "Ah, thanks, I had missed this somehow :)
", "rawMarkdown": "Ah, thanks, I had missed this somehow :)" } ] }, { "id": 2087642, "postDate": "2023-01-05T18:23:27.833Z", "content": "2 - train anywhere you want - upload model and predict - this is very common in kaggle challenges. But often a pain to get it to work :)
", "rawMarkdown": "2 - train anywhere you want - upload model and predict - this is very common in kaggle challenges. But often a pain to get it to work :)", "votes": 1, "replies": [ { "id": 2088659, "postDate": "2023-01-06T14:26:12.180Z", "content": "thanks for all your answers PC Jimmy, that's very helpful! :)
", "rawMarkdown": "thanks for all your answers PC Jimmy, that's very helpful! :)" } ] }, { "id": 2087646, "postDate": "2023-01-05T18:26:51Z", "content": "1 - to install packages using pip - the package needs to be in a data set that you have added. (the wheel file) It's a bit tricky to do this the first time so start simple with a single wheel in a data set you created.
", "rawMarkdown": "1 - to install packages using pip - the package needs to be in a data set that you have added. (the wheel file) It's a bit tricky to do this the first time so start simple with a single wheel in a data set you created. " } ] }, { "id": 2072764, "postDate": "2022-12-22T12:21:25.100Z", "content": "Hi, beginner question here.
\nI see others publishing datasets of converted PNGs in this competition. Is it ok to use these datasets in submissions? Does it invalidate the entry?
\nI think the alternative would be to include all image processing from DICOM to PNG in the submission code itself.
\nThanks.
", "rawMarkdown": "Hi, beginner question here. \n\nI see others publishing datasets of converted PNGs in this competition. Is it ok to use these datasets in submissions? Does it invalidate the entry? \n\nI think the alternative would be to include all image processing from DICOM to PNG in the submission code itself.\n\nThanks.", "replies": [ { "id": 2087652, "postDate": "2023-01-05T18:28:07.773Z", "content": "converted PNGs will NOT work in submission - because we do not have access to the test dicom files before hand to create them. In the submission you will need to read and process the dicom files as part of the script.
", "rawMarkdown": "converted PNGs will NOT work in submission - because we do not have access to the test dicom files before hand to create them. In the submission you will need to read and process the dicom files as part of the script." } ] }, { "id": 2062738, "postDate": "2022-12-12T10:58:42.567Z", "content": "Dear all
\nQuite new in Kaggle and Deep learning.
\nI have a question: what is the aim of this competition ?
\nIs it to detect a cancer from imagery or is it do diagnose (False Postive or False Negative or True Positve or True Negative) ?
\nAs it is mentioned in the headline \"It is important to identify cases of cancer for obvious reasons, but false positives also have downsides for patients.\"
Thanks for the tips!
", "rawMarkdown": "Thanks for the tips!" } ], "comments": [ { "id": 2085792, "author_name": "Tal Wac", "author_url": "", "post_date": "2023-01-04T11:57:52.937000", "content": "Hi all :)
\nI want to ask a question about downloading data.
\nWhen clicking **Data ** under the ribbon that include :
\nOverview Data Code Discussion Leaderboard Rules Team there is option/ tab that called
\n** Download All** which include 324.72 GB.
\nI do not have so much free space on the laptop I work on.
\nWanted to ask if there is an option to download part of the data?
I'm not sure if related, there is a link to Kaggle API installation and documentation, However I got lost there with unfamiliar terminology, and was sure if it's related to data downloading.
\nKindly your advice or guidance
\nTal
Hi Mohamed,
\nThank you for the detailed answer, much appeiciated!
I followed your steps and when I run the first block of your code in my Jupyter notebook I receive this error:
\nFileNotFoundError: [Errno 2] No such file or directory: 'train.csv'
Do I need to download this file 'train.csv' from somewhere?
\n(because the folder where I run my Jupyter notebook - \"C:\\Users.kaggle\\\" where kaggle.json is)
kindly your advice or guidance
\nTal
thank you very much its much clearer now :)
", "votes": 0, "replies": [] }, { "id": 2116413, "author_name": "Tal Wac", "author_url": "", "post_date": "2023-01-26T13:24:51.863000", "content": "Hi Mohamed,
\nI have another question with your permission.
\nthe code you wrote in part 3
\n(3) Iterate over the subset and **download every** image separately. My code:
however the first line in the code (which I rum from the jupyter note book, is this correct?)
\ndownloaded_images = [filename for filename in os.listdir() if '.dcm' in filename]
\nseems to assume I already have .dcm files in my folder (which I do not).
Can you please elaborate - how do I get .dcm files in my folder?
\nor is there something else that I should do?
\nThank you in advance
thank you Mohamed 🙏
", "votes": 0, "replies": [] } ] } ] }, { "id": 2534655, "author_name": "KUNAL SOOD", "author_url": "", "post_date": "2023-11-22T19:20:05.307000", "content": "hello everyone I want to work on this competition though its already over but data size if 372 gb i don't have much of space on my laptop is there any option to download part of the data?
", "votes": 0, "replies": [] }, { "id": 2165020, "author_name": "Ifeanyichukwu Nwobodo", "author_url": "", "post_date": "2023-03-01T23:48:47.210000", "content": "Any idea on how to incorporate Kaggle newly launched models in while training your model?
\nPreferably using fast ai. Thank you.
Hi,
\nMy question is about the submission and how the evaluation is done.
I'm quit lost in the workflow.
\nThe test data has only 4 files.
\nSo I do not understand where and at what stage all other 8000 patients (from the hidden test set) being added to the model for evaluation?
\nHow do I add all other 8000 patients with their images to the submission.csv file?
(I know I'm missing something not sure what)
\nkindly your help \\ guidance
", "votes": 0, "replies": [ { "id": 2132226, "author_name": "Tal Wac", "author_url": "", "post_date": "2023-02-06T17:12:55.417000", "content": "found the answer here
\nhttps://www.kaggle.com/competitions/rsna-breast-cancer-detection/discussion/382039
Hi there,
\nThis is a newbie kaggle question.
\nI'm running into issues due to not having internet access when making a competition submission.
\nTo this end, I'm looking for some tips on how to load a pretrained resnet50 model in the form of a .pth file.
\nI've copied it over to the appropriate checkpoint folder from /kaggle/input using the code below.
\nif not os.path.exists('/root/.cache/torch/hub/checkpoints/'):
\n os.makedirs('/root/.cache/torch/hub/checkpoints/')
\n!cp '/kaggle/input/resnet50-with-pytorch/resnet50.pth' '/root/.cache/torch/hub/checkpoints/resnet50.pth'
I've also defined a model class, however, I get the following error when I try to instantiate my model class.
\nNameError: name 'resnet50' is not defined
\nAny help or pointers would be greatly appreciated.
\nThank you,
\nBernie
could you post the first lines of your model class and the way you try to load your weights?
", "votes": 0, "replies": [] } ] }, { "id": 2087605, "author_name": "ellagale", "author_url": "", "post_date": "2023-01-05T17:53:54.597000", "content": "Hi All. I've used kaggle before, but years ago, and it seems that the set-up has changed. Please could you help asnwer my list of possibly stupid questions:
\nHow do we use open source software available on pip/conda? I've seen some notebooks installing packages from pip, when I try to install something I get : [Errno -3] Temporary failure in name resolution')', is there something else I am supposed to do? - i think I found a way to do this, using something called a wheel file, will update if it works
Can we train a model, save it, and then upload the model and use that to predict on the test data or do we need to submit a notebook that trains the model and then predicts on the test data as the submission?
I saw some mention of a time limit in the discussions, but there's nothing in the rules about this. Is there a time limit for training models or predicting on an image?
Thanks in advance for any help here.
", "votes": 0, "replies": [ { "id": 2087638, "author_name": "PC Jimmmy", "author_url": "", "post_date": "2023-01-05T18:21:36.670000", "content": "https://www.kaggle.com/competitions/rsna-breast-cancer-detection/overview/code-requirements
", "votes": 0, "replies": [ { "id": 2088660, "author_name": "ellagale", "author_url": "", "post_date": "2023-01-06T14:26:26.967000", "content": "Ah, thanks, I had missed this somehow :)
", "votes": 0, "replies": [] } ] }, { "id": 2087642, "author_name": "PC Jimmmy", "author_url": "", "post_date": "2023-01-05T18:23:27.833000", "content": "2 - train anywhere you want - upload model and predict - this is very common in kaggle challenges. But often a pain to get it to work :)
", "votes": 1, "replies": [ { "id": 2088659, "author_name": "ellagale", "author_url": "", "post_date": "2023-01-06T14:26:12.180000", "content": "thanks for all your answers PC Jimmy, that's very helpful! :)
", "votes": 0, "replies": [] } ] }, { "id": 2087646, "author_name": "PC Jimmmy", "author_url": "", "post_date": "2023-01-05T18:26:51", "content": "1 - to install packages using pip - the package needs to be in a data set that you have added. (the wheel file) It's a bit tricky to do this the first time so start simple with a single wheel in a data set you created.
", "votes": 0, "replies": [] } ] }, { "id": 2072764, "author_name": "frnkc", "author_url": "", "post_date": "2022-12-22T12:21:25.100000", "content": "Hi, beginner question here.
\nI see others publishing datasets of converted PNGs in this competition. Is it ok to use these datasets in submissions? Does it invalidate the entry?
\nI think the alternative would be to include all image processing from DICOM to PNG in the submission code itself.
\nThanks.
", "votes": 0, "replies": [ { "id": 2087652, "author_name": "PC Jimmmy", "author_url": "", "post_date": "2023-01-05T18:28:07.773000", "content": "converted PNGs will NOT work in submission - because we do not have access to the test dicom files before hand to create them. In the submission you will need to read and process the dicom files as part of the script.
", "votes": 0, "replies": [] } ] }, { "id": 2062738, "author_name": "Thibaut Jacquin", "author_url": "", "post_date": "2022-12-12T10:58:42.567000", "content": "Dear all
\nQuite new in Kaggle and Deep learning.
\nI have a question: what is the aim of this competition ?
\nIs it to detect a cancer from imagery or is it do diagnose (False Postive or False Negative or True Positve or True Negative) ?
\nAs it is mentioned in the headline \"It is important to identify cases of cancer for obvious reasons, but false positives also have downsides for patients.\"
Thanks for the tips!
", "votes": 0, "replies": [] } ], "raw_markdown_by_id": { "2047812": "New to machine learning and data science? No question is too basic or too simple. Feel free to start your own thread, or use this thread as a place to post any first-timer clarifying questions for the Kaggle community to help you with!\n\nIf you would consider yourself a beginner but don't know where to get started, let other Kagglers help you take your first steps here!\n\nNew to Kaggle? Take a look at a few videos Dr. Rachael Tatman has put together to learn a bit more about [site etiquette](https://www.youtube.com/watch?v=aIus8si_Et0), or [Kaggle lingo](https://www.youtube.com/watch?v=sEJHyuWKd-s).\n\nRemember: Kaggle is for everyone. Whether you're teaming up or sharing tips in the competition forum, we expect everyone to follow our [Kaggle community guidelines](https://www.kaggle.com/community-guidelines).\n\nA tip on sharing content - Kaggle is a collaborative community, whereby sharing techniques, starter notebooks, and ideas in the discussion forums are highly encouraged throughout the competition. However, as the competition draws closer to the final deadline it's customary to keep high-scoring notebooks withheld until after the competition has concluded. This maintains the spirit of the competition, while also allowing individuals to submit their own creative work without jeopardy of a higher-scoring notebook being available for an automatic higher rank (through copy/submit). We disable publishing of public notebooks within the final week of the competition, but encourage you to use your best judgment prior to that the deadline.\n\n \n\nHappy Modeling!\n", "2085792": "Hi all :) \nI want to ask a question about downloading data.\nWhen clicking **Data ** under the ribbon that include :\n*Overview Data Code Discussion Leaderboard Rules Team* there is option/ tab that called \n** Download All** which include 324.72 GB. \nI do not have so much free space on the laptop I work on. \nWanted to ask if there is an option to download part of the data? \n\nI'm not sure if related, there is a link to Kaggle API installation and documentation, However I got lost there with unfamiliar terminology, and was sure if it's related to data downloading. \nKindly your advice or guidance \nTal ", "2534655": "hello everyone I want to work on this competition though its already over but data size if 372 gb i don't have much of space on my laptop is there any option to download part of the data?\n", "2165020": "Any idea on how to incorporate Kaggle newly launched models in while training your model? \nPreferably using fast ai. Thank you.", "2132209": "Hi, \nMy question is about the submission and how the evaluation is done.\n\nI'm quit lost in the workflow. \n\nThe test data has only 4 files. \n\nSo I do not understand where and at what stage all other 8000 patients (from the hidden test set) being added to the model for evaluation?\nHow do I add all other 8000 patients with their images to the submission.csv file?\n\n(I know I'm missing something not sure what)\n\nkindly your help \\ guidance \n \n", "2111117": "Hi there, \n\nThis is a newbie kaggle question.\n\nI'm running into issues due to not having internet access when making a competition submission.\n\nTo this end, I'm looking for some tips on how to load a pretrained resnet50 model in the form of a .pth file.\n\nI've copied it over to the appropriate checkpoint folder from /kaggle/input using the code below.\n\nif not os.path.exists('/root/.cache/torch/hub/checkpoints/'):\n os.makedirs('/root/.cache/torch/hub/checkpoints/')\n!cp '/kaggle/input/resnet50-with-pytorch/resnet50.pth' '/root/.cache/torch/hub/checkpoints/resnet50.pth'\n\nI've also defined a model class, however, I get the following error when I try to instantiate my model class. \n\nNameError: name 'resnet50' is not defined\n\nAny help or pointers would be greatly appreciated. \n\nThank you,\nBernie", "2087605": "Hi All. I've used kaggle before, but years ago, and it seems that the set-up has changed. Please could you help asnwer my list of possibly stupid questions:\n\n1. How do we use open source software available on pip/conda? I've seen some notebooks installing packages from pip, when I try to install something I get : [Errno -3] Temporary failure in name resolution')', is there something else I am supposed to do? - i think I found a way to do this, using something called a wheel file, will update if it works\n\n2. Can we train a model, save it, and then upload the model and use that to predict on the test data or do we need to submit a notebook that trains the model and then predicts on the test data as the submission?\n\n3. I saw some mention of a time limit in the discussions, but there's nothing in the rules about this. Is there a time limit for training models or predicting on an image? \n\nThanks in advance for any help here.", "2072764": "Hi, beginner question here. \n\nI see others publishing datasets of converted PNGs in this competition. Is it ok to use these datasets in submissions? Does it invalidate the entry? \n\nI think the alternative would be to include all image processing from DICOM to PNG in the submission code itself.\n\nThanks.", "2062738": "Dear all\nQuite new in Kaggle and Deep learning.\nI have a question: what is the aim of this competition ?\nIs it to detect a cancer from imagery or is it do diagnose (False Postive or False Negative or True Positve or True Negative) ?\nAs it is mentioned in the headline \"It is important to identify cases of cancer for obvious reasons, but false positives also have downsides for patients.\"", "2157807": "Thanks for the tips!" } }