Data and example from Fernando Tussell, "Kalman Filtering in R", Journal of Statistical Software, http://www.jstatsoft.org/v39/i02/paper.

The data are three time series of exchange rates of BEF (Belgian Franc), CHF (Swiss Franc), and DEM (German Deutchmark)

Two models are estimated.

• fx_equicorr.stan: local level models for each series with a common measurement correlation.
• fx_factor.stan: Single factor model with local level factor.

Both models use the data in fx.data.R as input.

Example used in Fernando Tussell, "Kalman Filtering in R", Journal of Statistical Software, http://www.jstatsoft.org/v39/i02/paper, Durbin and Koopman (2001), and the papers in Journal of Statistical Software Vol 41, a special issue on state space models, http://www.jstatsoft.org/v41.

nile.stan is the model, and nile.data.R is the input data.