A column labeled “duration” may contain seconds, minutes, or a formatted piece of text. The meaning can seem obvious to its creator and become uncertain when the file is shared or reopened later.
Document the unit and keep the representation consistent. Distinguish a missing value from a zero, and state whether a timestamp carries a time zone. These small decisions matter when another tool reads the table.
Export a sample and inspect it outside the original application. Check that the values still have the intended meaning and that a reader can find their definitions. Portable data needs portable context as much as a widely readable file format.
Picture this situation.
Imagine exporting measurements to a colleague. Including the unit in the column description can preserve information that the original application supplied visually.
A second way to look.
Show a reader how to question the result. A date, an explanation of coverage, or a visible range can make a chart more useful than extra precision.
- Document units and representations.
- Define missing values separately from zero.
- Inspect an export outside the original tool.
Follow a related question
Distinguish model training from inference.
Training and use are different stagesCompare the same subject under different light.
Light becomes a digital pictureKeep learning
Related background to continue exploring this subject.
W3C: a primer for tabular data RFC Editor: the CSV format