Apologies, to be asking a dumb question.
\nI would like to know what's the difference between train.csv and train_features.csv files?
I am coming probably late into the game, as I joined with a friend of mine, just last night.
\nAlso, is there a document or something of that nature where in we can go read up about the data and its characteristics etc.,
", "messages": [ { "id": "1792938", "postDate": "05/17/2022 12:34:06", "content": "Apologies, to be asking a dumb question.
\nI would like to know what's the difference between train.csv and train_features.csv files?
I am coming probably late into the game, as I joined with a friend of mine, just last night.
\nAlso, is there a document or something of that nature where in we can go read up about the data and its characteristics etc.,
", "rawMarkdown": "Apologies, to be asking a dumb question. \nI would like to know what's the difference between **train.csv** and ***train_features.csv*** files?\n\nI am coming probably late into the game, as I joined with a friend of mine, just last night. \n\nAlso, is there a document or something of that nature where in we can go read up about the data and its characteristics etc.,", "votes": null }, { "id": "1793158", "postDate": "05/17/2022 16:30:33", "content": "@pankajashankar, First thing first, this competition is an extension of ASHRAE - Great Energy Predictor III. The provided dataset is the extended version of the same dataset used in the ASHRAE - Great Energy Predictor III competition after carefully annotating each data point of electricity meters from approximately 400 commercial buildings. So, please have a quick look into that dataset.
\nTo answer your question, as mentioned in the Data section, train_features.csv will have additional columns (including all columns in train.csv) that can be used as features to train your model. These features are extracted based on the metadata (building attributes and weather) from the ASHRAE - Great Energy Predictor III. You may check out this code to understand how those features are extracted. We provide this train_features.csv to help the participants focus on the modelling part rather redoing feature extraction.
\nI hope this helps!
", "rawMarkdown": "pankajashankar, First thing first, this competition is an extension of [ASHRAE - Great Energy Predictor III](https://www.kaggle.com/competitions/ashrae-energy-prediction/data). The provided dataset is the extended version of the same dataset used in the ASHRAE - Great Energy Predictor III competition after carefully annotating each data point of electricity meters from approximately 400 commercial buildings. So, please have a quick look into that [dataset](https://www.kaggle.com/competitions/ashrae-energy-prediction/data).\n\nTo answer your question, as mentioned in the [Data](https://www.kaggle.com/competitions/energy-anomaly-detection/data) section, **train_features.csv** will have additional columns (including all columns in **train.csv**) that can be used as features to train your model. These features are extracted based on the metadata (building attributes and weather) from the [ASHRAE - Great Energy Predictor III](https://www.kaggle.com/competitions/ashrae-energy-prediction/data). You may check out [this code](https://github.com/buds-lab/ashrae-great-energy-predictor-3-solution-analysis/blob/master/solutions/rank-1/scripts/02_preprocess_data.py) to understand how those features are extracted. We provide this **train_features.csv** to help the participants focus on the modelling part rather redoing feature extraction. \n\nI hope this helps!", "votes": null }, { "id": "1793271", "postDate": "05/17/2022 18:26:56", "content": "Thanks @samy101 for your clarification. I will look at the link that you mentioned to understand how the features were extracted.
", "rawMarkdown": "Thanks @samy101 for your clarification. I will look at the link that you mentioned to understand how the features were extracted.", "votes": null } ], "comments": [ { "id": 1793158, "author_name": "samy101", "author_url": "", "post_date": "05/17/2022 16:30:33", "content": "@pankajashankar, First thing first, this competition is an extension of ASHRAE - Great Energy Predictor III. The provided dataset is the extended version of the same dataset used in the ASHRAE - Great Energy Predictor III competition after carefully annotating each data point of electricity meters from approximately 400 commercial buildings. So, please have a quick look into that dataset.
\nTo answer your question, as mentioned in the Data section, train_features.csv will have additional columns (including all columns in train.csv) that can be used as features to train your model. These features are extracted based on the metadata (building attributes and weather) from the ASHRAE - Great Energy Predictor III. You may check out this code to understand how those features are extracted. We provide this train_features.csv to help the participants focus on the modelling part rather redoing feature extraction.
\nI hope this helps!
", "votes": null, "replies": [] }, { "id": 1793271, "author_name": "pankajashankar", "author_url": "", "post_date": "05/17/2022 18:26:56", "content": "Thanks @samy101 for your clarification. I will look at the link that you mentioned to understand how the features were extracted.
", "votes": null, "replies": [] } ], "raw_markdown_by_id": { "1792938": "Apologies, to be asking a dumb question. \nI would like to know what's the difference between **train.csv** and ***train_features.csv*** files?\n\nI am coming probably late into the game, as I joined with a friend of mine, just last night. \n\nAlso, is there a document or something of that nature where in we can go read up about the data and its characteristics etc.,", "1793158": "pankajashankar, First thing first, this competition is an extension of [ASHRAE - Great Energy Predictor III](https://www.kaggle.com/competitions/ashrae-energy-prediction/data). The provided dataset is the extended version of the same dataset used in the ASHRAE - Great Energy Predictor III competition after carefully annotating each data point of electricity meters from approximately 400 commercial buildings. So, please have a quick look into that [dataset](https://www.kaggle.com/competitions/ashrae-energy-prediction/data).\n\nTo answer your question, as mentioned in the [Data](https://www.kaggle.com/competitions/energy-anomaly-detection/data) section, **train_features.csv** will have additional columns (including all columns in **train.csv**) that can be used as features to train your model. These features are extracted based on the metadata (building attributes and weather) from the [ASHRAE - Great Energy Predictor III](https://www.kaggle.com/competitions/ashrae-energy-prediction/data). You may check out [this code](https://github.com/buds-lab/ashrae-great-energy-predictor-3-solution-analysis/blob/master/solutions/rank-1/scripts/02_preprocess_data.py) to understand how those features are extracted. We provide this **train_features.csv** to help the participants focus on the modelling part rather redoing feature extraction. \n\nI hope this helps!", "1793271": "Thanks @samy101 for your clarification. I will look at the link that you mentioned to understand how the features were extracted." }, "source": "meta" }