OK. I misunderstood what ARIMASim is good for.
In general: I would like to analyze time series with not equally spaced time data. Pattern components to be identified expected to have trend and seasonality (according to these descriptions: http://www.statsoft.com/textbook/time-series-analysis/) superposed with some noise (error). Some of the components predictable in advance, some of them not.
In particular: I’m attempting to use this library for the above problems.
I see various TimeSeries and their concrete classes, obvious holders of trend data, are they?
An ARIMA model has some parameters that drive its operation, I need to choose and set them.
What I cant see is how time series data is inputed to an ARIMA instance (or how to apply it on a data series).
It would be very didactic to see a complete non-trivial example of such.