{ "fetched_at": "2026-10-19T21:09:19Z", "source_url": "https://www.kaggle.com/competitions/dat-a-thon23/data", "pages": [ { "name": "data-description", "content": "\n## About the dataset\n\nThis dataset consists of a `train.csv` file which contains the unique id `id` for each row, 6 `category` columns with suffix A to F, 9 `feature` columns with suffix A to I, 10 `composition` columns with suffix from A to J, `unit` column which is the unit of measurement of the result value and finally a `result` columns which is the measure of the quality of water as a numerical value to be predicted.\n\n1. The `category` columns are various categorical features for a data point such as country of data collection, the site from which the data is collected, media of sample, etc.\n\n2. The `feature` columns are the various demographic features that affect the pollution of water in a particular region such as population density, GDP, droughts in a region, literacy rate of students in a region, etc.\n\n3. The `composition` columns are the compositions of various elements like paper, plastic wastes, cardboard, etc. in water.\n\n4. The `unit` value is the unit of measurement in which the result value is measured.\n\n5. The `result` value is a floating number that expresses the quality of water based on the various factors provided in the dataset.\n\n\n### FAQ\n\n**What files do I need?** \n\nYou are required to use the `train.csv` file to read the tabular dataset provided and train your algorithms to predict the numerical value in `result` column in `train.csv` file.\n \n**What am I predicting?**\n\nYou would be training your model based on the `train.csv` (training dataset) and then would be predicting the `result` value for each row in `test.csv` to create a submission. \n\n## Files\n\n* **train.csv** - the provided training data set.\n* **test.csv** - the provided testing data set.\n* **sample_submission.csv** - a sample submission file in the correct format.\n\n## Columns\n\n- `id`: The unique ID for each row.\n- `categoryA` - `categoryF`: 6 category columns with suffix A to F.\n- `featureA` - `featureI`: 9 feature columns with suffix A to I.\n- `compositionA` - `compositionJ`: 10 composition columns with suffix A to J.\n- `unit`: The unit of measurement for the result values.\n- `result`: The measure for water quality (target variable).\n\nThe training dataset provided could be read using the `read_csv()` function in the pandas module.", "isPublished": false, "hasIsPublished": false, "postTitle": "", "hasPostTitle": false, "mimeType": "", "hasMimeType": false }, { "name": "Description", "content": "## About the competition\nThe numbers are concerning, given that proper hand hygiene is one of the most effective measures to halt the spread of COVID-19 and other pathogens. Simple handwashing not only protects individuals from contracting the disease but also prevents transmission to others.\n\nHowever, it is alarming to note that 2.2 billion people worldwide lack access to safe water at home, and an additional 1.37 billion people lack handwashing facilities at home. Furthermore, nearly two billion people worldwide rely on healthcare facilities that lack basic water services.\n\nClimate change, population growth, and pollution are threatening the world's water resources. As the global population continues to expand, the challenge of accessing sufficient water while preserving aquatic ecosystems' integrity persists. The Pacific Institute collaborates with stakeholders worldwide to address water resource issues and ensure that communities and nature have the water they need to thrive presently and in the future.\n\nUnderstanding water sanitation and ensuring water cleanliness is crucial in both rural and urban areas. One way to achieve this is by assessing the quality of the water we consume daily.\n\nThe objective of this competition is to train a machine learning model using the provided water quality data in the training file and use it to predict the quality estimation result for the test dataset.\n\nFor further information on the competition, including instructions on submitting predictions, please refer to Kaggle's competition documentation available at the following link: https://www.kaggle.com/docs/competitions.\n\n## About the data set\nThe dataset provided in the `train.csv` consists of the following features:\n- `id`: The unique ID for each row.\n- `categoryA` - `categoryF`: 6 category columns with suffix A to F.\n- `featureA` - `featureI`: 9 feature columns with suffix A to I.\n- `compositionA` - `compositionJ`: 10 composition columns with suffix A to J.\n- `unit`: The unit of measurement for the result values.\n- `result`: The measure for water quality (target variable).\n\nThe datasets provided could be read using the `read_csv()` function in the pandas module.\n```\n# code to read the dataset\nimport pandas\ndf = pandas.read_csv(\"train.csv\")\n```\n\n## Acknowledgements\n\nWe thank European Environment Agency and The World Bank for providing this dataset.", "isPublished": false, "hasIsPublished": false, "postTitle": "", "hasPostTitle": false, "mimeType": "", "hasMimeType": false }, { "name": "Evaluation", "content": "Evaluation Metric\n---\nThe evaluation metric for this competition is **Root Mean Squared Logarithmic Error (RMSLE)**.\n\nThe RMSLE is calculated as\n\n\"RMSLE\n\nwhere:\n\\\\(n\\\\) is the total number of observations in the (public/private) data set,\n\\\\(xi\\\\) is your prediction of target, and \\\\(yi\\\\) is the actual target for \\\\(i\\\\).\n\\\\(\\log(x)\\\\) is the natural logarithm of \\\\(x\\\\)\n\n\n## Submission Format\n\nFor every `id` in the dataset, the submission file should contain two columns: `id` which is the unique id for each data point from the testing dataset, and `result` which is the water quality measurement factor. The second column should be a space-delimited string value.\n\nThe file should contain a header and have the following format:\n```\nid,result\n36c451925f31fbc7d248db938b9028711491a42c4e2524b3ca22a921169f06b8,0.0\n```", "isPublished": false, "hasIsPublished": false, "postTitle": "", "hasPostTitle": false, "mimeType": "", "hasMimeType": false }, { "name": "rules", "content": "Event Guidelines : \n\nPlease follow the guidelines I've included below to prepare the deliverable.\n--> The teams need to accept the invite by entering all the required details at Kaggle Invite Link. The Submission must be done before SampleDate to avoid disqualification of your entry.\n--> Use of Comments to explain the approach is preferred in the Notebook. Please make sure that the final .pynb file should be able to run on the Kaggle Platform. Also, any import necessary according to the platform must be present in the notebook.\nPresenters are welcome to ideate, animate & be creative in their submissions.\n--> Timings: The competition is open for 36 hours to check submissions, accuracy etc., and the final result be declared within 15 days of the submission date.\n--> Late submissions: Submissions made after the competition deadline will not be accepted unless extenuating circumstances in which the deadline is to be extended. All late submission requests must be emailed to ylp@growthjockey.com.\n--> This Kaggle Competition has a maximum limit of 20 submissions per day. All days are counted as in UTC Format.\n--> Evaluation criteria: Participants will be evaluated based on the evaluation metric specified in the competition rules.\n--> Team size: Participants may work individually or as part of a team, as specified in the competition rules. The maximum team size is 3.\n--> Prizes: Prizes will be awarded to the winners as specified in the competition rules. \n--> Disqualification: Participants may be disqualified if they violate any of the competition rules or engage in any unethical behavior such as plagiarism of code.\n\nPlease make sure that you prepare the deliverable & upload a python notebook file on the following link: https://form.typeform.com/to/uI2gc6Eu.\n\n\n", "isPublished": false, "hasIsPublished": false, "postTitle": "", "hasPostTitle": false, "mimeType": "", "hasMimeType": false } ] }