Time-Series Data

Time-series data is any sequence of values recorded against timestamps, from tick-by-tick prices to daily foot-traffic counts. Its defining feature is that order and spacing carry meaning: the same values in a different sequence describe a different world. Series can be regularly sampled (one value per minute) or irregular (a value whenever an event fires), and turning one into the other by resampling is a routine but error-prone step.

Working with time series demands discipline about time itself. Mixing time zones, aligning series that update on different schedules, and - most dangerously - letting a future value leak into a calculation meant to run in the past are the classic mistakes. That last one, look-ahead bias, is why serious analysis pairs time-series data with point-in-time history that preserves exactly what was known at each moment.