Datapiste

Datapiste is a term used in data science and information management to denote a single observed unit within a dataset. It is often treated as equivalent to a data point, a row in a table, or a record, but some writers emphasize its role in tracking provenance and transformations through a data pipeline.

Etymology and usage: The word combines the English data with the French piste, meaning track or trace. It appears mainly in French-language documentation and in multilingual contexts, but it is not standardized and may be used inconsistently across disciplines.

Structure and examples: A datapiste typically comprises fields such as a timestamp, a source identifier, a measured value, and optional metadata like units and quality flags. In time-series data, each datapiste represents a timestamped observation; in a relational table it corresponds to a row, and in a wide dataset it may span multiple columns.

Relation to related concepts: Datapiste is closely related to data point, data record, and observation. It is sometimes preferred when emphasis is on traceability through transformations, lineage tracking, or auditability, but many practitioners simply use data point or record.

Related terms include data point, data record, time series, and data lineage.