Time series analysis and forecast, Trends of different resolution

Examples: Time series analysis

Trends of different resolution

The ‘Caterpillar’-SSA is a convenient technique to extract trends of different resolution. Natural decomposition of the time series is constructed on the base of the unique parameter (the window length). Grouping different subsets of the decomposition components, you can obtain both the tendency and the accurate trend.

Figure 1.1: Production: General tendency (rough trend)

Figure 1.2: Production: Accurate trend

Data: "Production"
Description: crude oil, lease condensate, and natural gas plant liquids production, monthly data from January 1973 to September 1999
Source: Economic Time Series Page
Parameters: window length = 120
Figure 1.1: the time series component reconstructed from the leading eigentriple (red line) and the initial time series
Figure 1.2: reconstruction from the leading 3 eigentriples (red line) and the initial time series

Previous example | Next example