The contest Data page says that train_metadata.csv has 5 columns: ebird_code, recodist, site, date, filename. However, this file actually has 14 columns. It has date and filename as described, but instead of site it has latitude, longitude; instead of recodist it has author; instead of ebird_code it has primary_label and secondary_labels; and then it also has type, scientific_name, common_name, license, rating, time and url. Please update the Data page.
The contest Data page says that train_metadata.csv has 5 columns: ebird_code, recodist, site, date, filename. However, this file actually has 14 columns. It has date and filename as described, but instead of site it has latitude, longitude; instead of recodist it has author; instead of ebird_code it has primary_label and secondary_labels; and then it also has type, scientific_name, common_name, license, rating, time and url. Please update the Data page.
Thanks for pointing this out! Seems to be confused with last year's description. We will update the \"data\" page. Good catch!
", "rawMarkdown": "Thanks for pointing this out! Seems to be confused with last year's description. We will update the \"data\" page. Good catch!", "votes": null }, { "id": "1264149", "postDate": "04/05/2021 23:44:38", "content": "Thanks for pointing this out. The issue has been resolved!
", "rawMarkdown": "Thanks for pointing this out. The issue has been resolved!", "votes": null }, { "id": "1264206", "postDate": "04/06/2021 01:43:35", "content": "It looks like the \"Data\" page has not been changed.
", "rawMarkdown": "It looks like the \"Data\" page has not been changed.", "votes": null }, { "id": "1264548", "postDate": "04/06/2021 08:39:32", "content": "It should be updated now.
", "rawMarkdown": "It should be updated now.", "votes": null }, { "id": "1264744", "postDate": "04/06/2021 12:06:43", "content": "OK, as for the data description in train_metadata.csv, I see that only the most directly relevant fields are listed. But, I'd like to know the exact meaning of 'secondary_labels', 'type', 'author' and 'rating'.
\nI guess they are:
\n secondary_labels: birds list of lower sound volume than the primary
\n type: scene of bird sound
\n author: the same as 'recodist', which column isn't there
\n rating: recoding quality: the higher the better
\nIs that right?
Yes, it is. I also added a few notes on all metadata fields here: www.kaggle.com/stefankahl/birdclef2021-exploring-the-data
", "rawMarkdown": "Yes, it is. I also added a few notes on all metadata fields here: www.kaggle.com/stefankahl/birdclef2021-exploring-the-data", "votes": null } ], "comments": [ { "id": 1262385, "author_name": "stefankahl", "author_url": "", "post_date": "04/04/2021 07:44:01", "content": "Thanks for pointing this out! Seems to be confused with last year's description. We will update the \"data\" page. Good catch!
", "votes": null, "replies": [] }, { "id": 1264149, "author_name": "holgerklinck", "author_url": "", "post_date": "04/05/2021 23:44:38", "content": "Thanks for pointing this out. The issue has been resolved!
", "votes": null, "replies": [ { "id": 1264206, "author_name": "nak326", "author_url": "", "post_date": "04/06/2021 01:43:35", "content": "It looks like the \"Data\" page has not been changed.
", "votes": null, "replies": [] }, { "id": 1264548, "author_name": "stefankahl", "author_url": "", "post_date": "04/06/2021 08:39:32", "content": "It should be updated now.
", "votes": null, "replies": [] }, { "id": 1264744, "author_name": "nak326", "author_url": "", "post_date": "04/06/2021 12:06:43", "content": "OK, as for the data description in train_metadata.csv, I see that only the most directly relevant fields are listed. But, I'd like to know the exact meaning of 'secondary_labels', 'type', 'author' and 'rating'.
\nI guess they are:
\n secondary_labels: birds list of lower sound volume than the primary
\n type: scene of bird sound
\n author: the same as 'recodist', which column isn't there
\n rating: recoding quality: the higher the better
\nIs that right?
Yes, it is. I also added a few notes on all metadata fields here: www.kaggle.com/stefankahl/birdclef2021-exploring-the-data
", "votes": null, "replies": [] } ] } ], "raw_markdown_by_id": { "1262297": "The contest Data page says that `train_metadata.csv` has 5 columns: `ebird_code`, `recodist`, `site`, `date`, `filename`. However, this file actually has 14 columns. It has `date` and `filename` as described, but instead of `site` it has `latitude`, `longitude`; instead of `recodist` it has `author`; instead of `ebird_code` it has `primary_label` and `secondary_labels`; and then it also has `type`, `scientific_name`, `common_name`, `license`, `rating`, `time` and `url`. Please update the Data page.", "1262385": "Thanks for pointing this out! Seems to be confused with last year's description. We will update the \"data\" page. Good catch!", "1264149": "Thanks for pointing this out. The issue has been resolved!", "1264206": "It looks like the \"Data\" page has not been changed.", "1264548": "It should be updated now.", "1264744": "OK, as for the data description in train_metadata.csv, I see that only the most directly relevant fields are listed. But, I'd like to know the exact meaning of 'secondary_labels', 'type', 'author' and 'rating'.\nI guess they are:\n secondary_labels: birds list of lower sound volume than the primary\n type: scene of bird sound\n author: the same as 'recodist', which column isn't there\n rating: recoding quality: the higher the better\nIs that right?", "1265138": "Yes, it is. I also added a few notes on all metadata fields here: www.kaggle.com/stefankahl/birdclef2021-exploring-the-data" }, "source": "meta" }