Periodic patterns in time series

Often, there are recurring trends in the data that suggest a periodic pattern, and simple plotting of the time series (like what we did in Chapter 2, Dissecting Data Using Mathematica) might not reveal such patterns. Besides, some datasets are inherently periodic. The daily Internet usage of a community can be represented as the hourly percentage of people using the Internet over a span of 24 hours. We can collect such data for a month and visualize any visible trend present in the data. So, what we will have is a percentage number between 0 and 1, recorded for each hour over 30 days. If we are given such a structured dataset, it might not be wise to simply plot the points over time. We would like to compare hourly activity side by side so that we can compare any difference on an hourly basis.

There is an easy solution if we want to stick to point plotting. We can use the Table function and the ;; operator to divide the data into 30 sets of equal length; this...