A common approach in tutorials is to jump right into manipulating a plotting time series by using one of R's many built-in time series data sets. Here, I'd like to explain how to actually create a time series object (an R object of type ts) like you would if you were to have to perform forecasting on your own data.
I made available some special data sets that we'll use in this and future chapters by putting on a repository on the internet.
Let's first start by downloading one of these data sets and peering into its contents:
> download.file("https://raw.githubusercontent.com/tonyfischetti/dawr/master/forecasting/school-supplies.csv", "./school-supplies.csv")
> school <- read.csv("./school-supplies.csv")
> head(school)
thedate hits
1 2004-01 20
2 2004-02 24
3 2004-03 19
4 2004-04 26
5 2004-05 25
6 2004-06 24
> start(schoolts)
[1] 2004 1
> end(schoolts)
[1] 2018 3This is a data set that contains the number of Google searches for school supplies...