Introducing different sources of time series data

In this section, we’ll introduce real-world sources of time series data that represent a wide range of possible characteristics of time series data.

Time series data contains, besides the physical dimension of time, measurements of a quantitative variable such as temperature, a financial KPI, or sales. In the raw data, the progress of this variable can be reported at regular or irregular intervals. In regular time series, the sampling frequency is predefined and constant, whereas, in irregular time series, the timestamps are generated based on the occurrences of random events.

For example, if we record the total sales at the end of each day, we produce a time series with regular, daily intervals. If we, instead, record the sales generated by each customer in a day, we produce a time series with irregular intervals and possibly multiple values for the same timestamp. However, both time series can be analyzed by time series...