Trend and seasonality components
If we consider the graphical analysis presented in the previous chapter to describe a Time Series, it is evident that there are different patterns that characterize the temporal dynamics of the data. We have already spoken, several times, in the book about trend or seasonality; however, it is necessary to go into more detail about these components in order to subsequently understand how to treat them within a forecasting model. Let’s start with trends.
Trend
Trend can be described as the direction in which the Time Series is running during a long period. So, in general, a trend exists when there is a long-term increase (that is, the upward trend) or decrease (that is, the downward trend) in some Time Series data. Typically, it’s difficult to formally define “trend” because its characteristics depend on the framework where the trend is analyzed. For instance, the formal definition of trend could be different depending...
