Seasonsfrom

Seasonsfrom is a term used in time-series analysis to describe the collection of sources or origins of seasonal variation observed in a data series. It is an analytic construct rather than a fixed statistical object, intended to capture the factors that generate recurring patterns at regular intervals, such as monthly, quarterly, or annual cycles, holidays, or climate-driven effects.

In decomposition frameworks, data are often split into trend, seasonal, and irregular components. Seasonsfrom refers to the set of potential drivers that contribute to the seasonal component, including calendar effects, institutional schedules, and exogenous variables. It helps researchers reason about why seasonality occurs, not merely that seasonality does occur.

Practically, seasonsfrom can be inferred by examining seasonal factors estimated by models like STL, classical decomposition, or state-space methods, and by linking the seasonal component to candidate sources. However, identifying explicit sources is challenging and often requires external data or domain knowledge.

It is distinct from seasonal adjustment, which aims to remove seasonal variation, whereas seasonsfrom emphasizes understanding its origins.

See also: seasonality; seasonal adjustment; decomposition; STL; X-13ARIMA-SEATS.