The use of descriptive statistics and data visualization tools plays a pivotal role in the seasonality analysis of time series data. As we saw in this chapter, there is a close relationship between the frequency of the series and the type of seasonal patterns. A series with a lower frequency (such as monthly or quarterly) would potentially have a single dominant seasonal pattern. On the other hand, if the series frequency is higher, the probability is that multiple seasonal patterns exist in the series. This, of course, should help you to determine which tools or approaches to use in the analysis process. Last but not least, in some instances, you should consider removing exogenous factors (such as the series trend) to get a clear picture of the seasonal patterns and to avoid misleading results.
In the next chapter, we will focus on the correlation analysis of time series...
