Types of pattern

Time series can have four types of pattern. These patterns affect the interpretation of the data and are used to help make forecasts of future values.


The exercise below shows a time series that has two of the above patterns. Use the pop-up menus under the time series plot to identify these patterns.

Note that if trend is present, it is linear trend. A cyclical pattern may appear to be a nonlinear trend in a short time series, so if you see curvature, it is due to cyclical pattern.

Repeat with a few different questions.