This is my submission for a coding task for a position as data scientist.
The task was to define a time series, to translate it to its symbolic representation using Symbolic Aggregate approXimation (SAX), and to compute the frequency of each SAX symbol. Furthermore, the task was to provide a graphical user interface to explore the impact of various parameters.
For my solution, I start by constructing an ARMA(1,1) time series with 100 periods. A graphical user interface then allows to choose frame size and alphabet size for a SAX representation, and to view the resulting frequency distribution of categories, as well as the time series itself.