Eyy, vertebrae virtuosos.Khaled here.
\nSpine anatomy? Greek to me. Kaggle setup? A maze.
\nCan't submit to the leaderboard. It's killin' me. I am struggling building a pipeline that take input and make a submission into the leaderboard. That is my main goal for now.
\nLookin' for a pipeline blueprint. Help a wiseguy out?
\nIf not, I'll muscle through solo.
\nNow, if you'll excuse me, I've got some vertebrae to crack haha 🤣
Eyy, vertebrae virtuosos.Khaled here.
\nSpine anatomy? Greek to me. Kaggle setup? A maze.
\nCan't submit to the leaderboard. It's killin' me. I am struggling building a pipeline that take input and make a submission into the leaderboard. That is my main goal for now.
\nLookin' for a pipeline blueprint. Help a wiseguy out?
\nIf not, I'll muscle through solo.
\nNow, if you'll excuse me, I've got some vertebrae to crack haha 🤣
Look under Code tab of this competition. Various people have given both training and inference notebooks. Example:
\n", "rawMarkdown": "Look under [Code](https://www.kaggle.com/competitions/rsna-2024-lumbar-spine-degenerative-classification/code) tab of this competition. Various people have given both training and inference notebooks. Example:\n- [RSNA2024 LSDC Training DenseNet](https://www.kaggle.com/code/hugowjd/rsna2024-lsdc-training-densenet)\n- [RSNA2024 LSDC DenseNet Submission](https://www.kaggle.com/code/hugowjd/rsna2024-lsdc-densenet-submission)", "votes": null }, { "id": "2949697", "postDate": "08/06/2024 22:09:50", "content": "Thanks for the heads up, @coderrkj
\n I appreciate your time
\nIll check them for sure
There a a lot of awesome sharings in Code. Here is a minimal submission script:
\n pandas pd\n\ndf = pd.read_csv(\n \n)\n\n\nCONDITIONS = [\n , \n , \n ,\n ,\n \n]\nLEVELS = [\n ,\n ,\n ,\n ,\n ,\n]\n\n random random\n numpy np\n\n ():\n \n [, , ]\n\nsubmission_df = []\n study_id df[].unique():\n \n condition CONDITIONS:\n level LEVELS:\n new_row = [\n .join([(study_id), condition, level])\n ]\n new_row += predict()\n submission_df.append(new_row)\nsubmission_df = pd.DataFrame(submission_df, columns=[, , , ])\nsubmission_df.to_csv(, index=)\nyou may replace predict() with your model inference code and pass everything you need like study_id, etc.
", "rawMarkdown": "There a a lot of awesome sharings in [Code](https://www.kaggle.com/competitions/rsna-2024-lumbar-spine-degenerative-classification/code). Here is a minimal submission script:\n```\nimport pandas as pd\n\ndf = pd.read_csv(\n '/kaggle/input/rsna-2024-lumbar-spine-degenerative-classification/test_series_descriptions.csv'\n)\n\n# need to predict every condiction_level (total 25) for every study\nCONDITIONS = [\n 'spinal_canal_stenosis', \n 'left_neural_foraminal_narrowing', \n 'right_neural_foraminal_narrowing',\n 'left_subarticular_stenosis',\n 'right_subarticular_stenosis'\n]\nLEVELS = [\n 'l1_l2',\n 'l2_l3',\n 'l3_l4',\n 'l4_l5',\n 'l5_s1',\n]\n\nfrom random import random\nimport numpy as np\n\ndef predict():\n # return [\"normal_mild\", \"moderate\", \"severe\"] scores for every studyid_condition_level\n return [0.6, 0.3, 0.1]\n\nsubmission_df = []\nfor study_id in df[\"study_id\"].unique():\n # depends on your model's output\n for condition in CONDITIONS:\n for level in LEVELS:\n new_row = [\n \"_\".join([str(study_id), condition, level])\n ]\n new_row += predict()\n submission_df.append(new_row)\nsubmission_df = pd.DataFrame(submission_df, columns=[\"row_id\", \"normal_mild\", \"moderate\", \"severe\"])\nsubmission_df.to_csv('submission.csv', index=False)\n```\nyou may replace predict() with your model inference code and pass everything you need like study_id, etc.", "votes": null } ], "comments": [ { "id": 2947298, "author_name": "coderrkj", "author_url": "", "post_date": "08/05/2024 05:50:34", "content": "Look under Code tab of this competition. Various people have given both training and inference notebooks. Example:
\n", "votes": null, "replies": [ { "id": 2949697, "author_name": "tinkerturbo", "author_url": "", "post_date": "08/06/2024 22:09:50", "content": "Thanks for the heads up, @coderrkj
\n I appreciate your time
\nIll check them for sure
There a a lot of awesome sharings in Code. Here is a minimal submission script:
\n pandas pd\n\ndf = pd.read_csv(\n \n)\n\n\nCONDITIONS = [\n , \n , \n ,\n ,\n \n]\nLEVELS = [\n ,\n ,\n ,\n ,\n ,\n]\n\n random random\n numpy np\n\n ():\n \n [, , ]\n\nsubmission_df = []\n study_id df[].unique():\n \n condition CONDITIONS:\n level LEVELS:\n new_row = [\n .join([(study_id), condition, level])\n ]\n new_row += predict()\n submission_df.append(new_row)\nsubmission_df = pd.DataFrame(submission_df, columns=[, , , ])\nsubmission_df.to_csv(, index=)\nyou may replace predict() with your model inference code and pass everything you need like study_id, etc.
", "votes": null, "replies": [] } ], "raw_markdown_by_id": { "2947053": "Eyy, vertebrae virtuosos.Khaled here.\nSpine anatomy? Greek to me. Kaggle setup? A maze.\nCan't submit to the leaderboard. It's killin' me. I am struggling building a pipeline that take input and make a submission into the leaderboard. That is my main goal for now. \nLookin' for a pipeline blueprint. Help a wiseguy out?\nIf not, I'll muscle through solo.\nNow, if you'll excuse me, I've got some vertebrae to crack haha 🤣", "2947298": "Look under [Code](https://www.kaggle.com/competitions/rsna-2024-lumbar-spine-degenerative-classification/code) tab of this competition. Various people have given both training and inference notebooks. Example:\n- [RSNA2024 LSDC Training DenseNet](https://www.kaggle.com/code/hugowjd/rsna2024-lsdc-training-densenet)\n- [RSNA2024 LSDC DenseNet Submission](https://www.kaggle.com/code/hugowjd/rsna2024-lsdc-densenet-submission)", "2949697": "Thanks for the heads up, @coderrkj \n I appreciate your time\nIll check them for sure", "2949886": "There a a lot of awesome sharings in [Code](https://www.kaggle.com/competitions/rsna-2024-lumbar-spine-degenerative-classification/code). Here is a minimal submission script:\n```\nimport pandas as pd\n\ndf = pd.read_csv(\n '/kaggle/input/rsna-2024-lumbar-spine-degenerative-classification/test_series_descriptions.csv'\n)\n\n# need to predict every condiction_level (total 25) for every study\nCONDITIONS = [\n 'spinal_canal_stenosis', \n 'left_neural_foraminal_narrowing', \n 'right_neural_foraminal_narrowing',\n 'left_subarticular_stenosis',\n 'right_subarticular_stenosis'\n]\nLEVELS = [\n 'l1_l2',\n 'l2_l3',\n 'l3_l4',\n 'l4_l5',\n 'l5_s1',\n]\n\nfrom random import random\nimport numpy as np\n\ndef predict():\n # return [\"normal_mild\", \"moderate\", \"severe\"] scores for every studyid_condition_level\n return [0.6, 0.3, 0.1]\n\nsubmission_df = []\nfor study_id in df[\"study_id\"].unique():\n # depends on your model's output\n for condition in CONDITIONS:\n for level in LEVELS:\n new_row = [\n \"_\".join([str(study_id), condition, level])\n ]\n new_row += predict()\n submission_df.append(new_row)\nsubmission_df = pd.DataFrame(submission_df, columns=[\"row_id\", \"normal_mild\", \"moderate\", \"severe\"])\nsubmission_df.to_csv('submission.csv', index=False)\n```\nyou may replace predict() with your model inference code and pass everything you need like study_id, etc." }, "source": "meta" }