scaletype

Scaletype is a term used in data visualization and plotting to describe the method by which data values are transformed and mapped to visual coordinates or display attributes. It defines how the input domain of data is converted into the output range, shaping spacing, labeling, and interpretation of axes. Different libraries may call it “scale type” or simply use a scale with a specific kind.

Common scaletypes include linear, where values are evenly spaced on the axis; logarithmic, which expands small ranges and compresses large ranges to handle data spanning multiple orders of magnitude; symlog, which behaves linearly near zero and logarithmically away from zero for data that crosses zero; logit, which is suitable for values between 0 and 1 (often used with probabilities); time scales, which map date or time values to positions with appropriate tick generation; and ordinal or categorical scales, which place discrete categories at evenly spaced positions. Some libraries also offer power or square-root scales for particular data relationships. In practice, color or size channels may adopt analogous scale concepts.

Choosing a scaletype depends on the data distribution and the visualization goal. Use linear for roughly uniform data, logarithmic for broad ranges or multiplicative relationships, time for temporal data, and ordinal for distinct categories. Consider how ticks should appear, whether zeros are meaningful, and whether interactive features require invertibility or clamping of values.

Implementation notes often include defining a domain (input data extent) and a range (visual span), enabling clamping to restrict outputs, supporting invert() for interactive zooming, and allowing automatic tick generation or manual control. Different plotting libraries implement scaletypes with various names but share the same core idea of transforming data to the visual plane. See also scales, axes, and tick generation.