Are there multiple ways of encoding zero masks?
\n\nI think according to the definition of RLE encoding, an image with zero pixels might be encoded \nminimally in two forms: an empty set \"\" or \"1 0\". Are both accepted as valid from a scoring perspective?
", "messages": [ { "id": "312982", "postDate": "04/12/2018 16:08:13", "content": "Are there multiple ways of encoding zero masks?
\n\nI think according to the definition of RLE encoding, an image with zero pixels might be encoded \nminimally in two forms: an empty set \"\" or \"1 0\". Are both accepted as valid from a scoring perspective?
", "rawMarkdown": "Are there multiple ways of encoding zero masks?\n\nI think according to the definition of RLE encoding, an image with zero pixels might be encoded \nminimally in two forms: an empty set \"\" or \"1 0\". Are both accepted as valid from a scoring perspective?", "votes": null }, { "id": "312999", "postDate": "04/12/2018 16:43:07", "content": "Nice question, let's see answer from organizers.
", "rawMarkdown": "Nice question, let's see answer from organizers.", "votes": null }, { "id": "313006", "postDate": "04/12/2018 16:50:46", "content": "I didn't see there was a duplicate question posted before mine: https://www.kaggle.com/c/data-science-bowl-2018/discussion/54359 @Heng has a probing result that says it is safe to go with empty.
", "rawMarkdown": "I didn't see there was a duplicate question posted before mine: https://www.kaggle.com/c/data-science-bowl-2018/discussion/54359 @Heng has a probing result that says it is safe to go with empty.", "votes": null }, { "id": "313007", "postDate": "04/12/2018 16:50:47", "content": "This does need to be cleared up by the organizers. Specifically what is the correct way to encode RLE for zero pixels, so that the scoring code will not penalize the submission. There have been suggestions of \"empty strings\" and \"1 0\". I myself have used \"0 0\" in my code but have not made a submission yet. A definitive answer would be great seeing as we have a limited amount of time left.
", "rawMarkdown": "This does need to be cleared up by the organizers. Specifically what is the correct way to encode RLE for zero pixels, so that the scoring code will not penalize the submission. There have been suggestions of \"empty strings\" and \"1 0\". I myself have used \"0 0\" in my code but have not made a submission yet. A definitive answer would be great seeing as we have a limited amount of time left.", "votes": null }, { "id": "313744", "postDate": "04/13/2018 19:24:09", "content": "Empty string works.
", "rawMarkdown": "Empty string works.", "votes": null }, { "id": "313747", "postDate": "04/13/2018 19:25:39", "content": "Thank you.
", "rawMarkdown": "Thank you.", "votes": null } ], "comments": [ { "id": 312999, "author_name": "maksimovka", "author_url": "", "post_date": "04/12/2018 16:43:07", "content": "Nice question, let's see answer from organizers.
", "votes": null, "replies": [ { "id": 313006, "author_name": "pliptor", "author_url": "", "post_date": "04/12/2018 16:50:46", "content": "I didn't see there was a duplicate question posted before mine: https://www.kaggle.com/c/data-science-bowl-2018/discussion/54359 @Heng has a probing result that says it is safe to go with empty.
", "votes": null, "replies": [] } ] }, { "id": 313007, "author_name": "robertkag", "author_url": "", "post_date": "04/12/2018 16:50:47", "content": "This does need to be cleared up by the organizers. Specifically what is the correct way to encode RLE for zero pixels, so that the scoring code will not penalize the submission. There have been suggestions of \"empty strings\" and \"1 0\". I myself have used \"0 0\" in my code but have not made a submission yet. A definitive answer would be great seeing as we have a limited amount of time left.
", "votes": null, "replies": [] }, { "id": 313744, "author_name": "wendykan", "author_url": "", "post_date": "04/13/2018 19:24:09", "content": "Empty string works.
", "votes": null, "replies": [ { "id": 313747, "author_name": "pliptor", "author_url": "", "post_date": "04/13/2018 19:25:39", "content": "Thank you.
", "votes": null, "replies": [] } ] } ], "raw_markdown_by_id": { "312982": "Are there multiple ways of encoding zero masks?\n\nI think according to the definition of RLE encoding, an image with zero pixels might be encoded \nminimally in two forms: an empty set \"\" or \"1 0\". Are both accepted as valid from a scoring perspective?", "312999": "Nice question, let's see answer from organizers.", "313006": "I didn't see there was a duplicate question posted before mine: https://www.kaggle.com/c/data-science-bowl-2018/discussion/54359 @Heng has a probing result that says it is safe to go with empty.", "313007": "This does need to be cleared up by the organizers. Specifically what is the correct way to encode RLE for zero pixels, so that the scoring code will not penalize the submission. There have been suggestions of \"empty strings\" and \"1 0\". I myself have used \"0 0\" in my code but have not made a submission yet. A definitive answer would be great seeing as we have a limited amount of time left.", "313744": "Empty string works.", "313747": "Thank you." }, "source": "meta" }