A dataset (or data set) is a collection of data.
Most commonly a dataset corresponds to the contents of a single database table, or a single statistical data matrix, where each column of the table represents a particular variable, and each row corresponds to a given member of the dataset in question. The dataset lists values for each of the variables, such as height and weight of an object, for each member of the dataset. Each value is known as a datum. The dataset may comprise data for one or more members, corresponding to the number of rows.
The term dataset may also be used more loosely, to refer to the data in a collection of closely related tables, corresponding to a particular experiment or event.
Several characteristics define a dataset's structure and properties. These include the number and types of the attributes or variables, and various statistical measures applicable to them, such as standard deviation and kurtosis.
In the simplest case, there is only one variable, and the dataset consists of a single column of values, often represented as a list. In spite of the name, such a univariate dataset is not a set in the usual mathematical sense, since a given value may occur multiple times. Usually the order does not matter, and then the collection of values may be considered a multiset rather than an (ordered) list[original research?].
The values may be numbers, such as real numbers or integers, for example representing a person's height in centimeters, but may also be nominal data (i.e., not consisting of numerical values), for example representing a person's ethnicity. More generally, values may be of any of the kinds described as a level of measurement. For each variable, the values are normally all of the same kind. However, there may also be missing values, which must be indicated in some way.
In statistics, datasets usually come from actual observations obtained by sampling a statistical population, and each row corresponds to the observations on one element of that population. Datasets may further be generated by algorithms for the purpose of testing certain kinds of software. Some modern statistical analysis software such as SPSS still present their data in the classical dataset fashion
Several classic datasets have been used extensively in the statistical literature:
- Iris flower data set - multivariate dataset introduced by Ronald Fisher (1936).
- Categorical data analysis - Datasets used in the book, An Introduction to Categorical Data Analysis, by Agresti are provided on-line by StatLib.
- Robust statistics - Datasets used in Robust Regression and Outlier Detection (Rousseeuw and Leroy, 1986). Provided on-line at the University of Cologne.
- Time series - Data used in Chatfield's book, The Analysis of Time Series, are provided on-line by StatLib.
- Extreme values - Data used in the book, An Introduction to the Statistical Modeling of Extreme Values are a snapshot of the data as it was provided on-line by Stuart Coles, the book's author.
- Bayesian Data Analysis - Data used in the book are provided on-line by Andrew Gelman, one of the book's authors.
- The Bupa liver data, used in several papers in the machine learning (data mining) literature.
- Anscombe's quartet Small dataset illustrating the importance of graphing the data to avoid statistical fallacies
- Datahub - A community-managed home for open datasets
- Research Pipeline - A wiki/website with links to datasets on many different topics.
- StatLib--Datasets Archive
- StatLib--JASA Data Archive
- UK Government Public Data
- GCMD - The Global Change Master Directory contains more than 20,000 descriptions of Earth science datasets and services covering all aspects of Earth and environmental sciences.