: The World Health Organization provides extensive data on alcohol consumption and health outcomes globally.
The drinks dataset proves that you don’t need millions of rows to ask interesting questions. Whether you’re learning to group data, build your first bar chart, or detect regional patterns, this humble collection of beer, wine, and spirits servings offers a refreshing start.
Predominantly Muslim-majority countries and those with strict alcohol laws report near-zero consumption: Afghanistan, Pakistan, Mauritania, and Bangladesh.
If you scrape a review dataset, remember that seasonal beers (e.g., Pumpkin Ale in October, Winter Stouts in December) will skew your ABV and style distribution if you do not account for the date of review.
If the five listed above do not fit your specific needs (for example, you need data on kombucha sales or energy drink market share), you will need to dig deeper. Do not just Google "drinks dataset"; try these specific sources:
Starbucks has made its menu nutritional data public, which serves as an excellent drinks dataset for regression and classification tasks.
A prime example often used in introductory data science is the "Alcohol Consumption by Country" dataset, which tracks per capita consumption of beer, wine, and spirits across different nations.