ftsatoo

Derived from R package ftsa.

Functions for visualizing, modeling, (generic) forecasting and hypothesis testing of functional time series.

library(ftsatoo)
library(ahead)

fit_obj <- ftsm(ElNino_ERSST_region_1and2, order=3, method = "M")
summary(fit_obj)
plot(fit_obj)

# Here's the novelty
(fcast <- forecast(object = fit_obj, h = 12, level=95, 
                    pimethod = "nonparametric",
                    FUN=forecast::thetaf))   
plot(fcast)                    

(fcast <- forecast(object = fit_obj, h = 12, level=95, 
                    pimethod = "nonparametric",
                    FUN=ahead::dynrmf))     
plot(fcast)                                      

# And here too (multivariate)
(fcast <- forecast(object = fit_obj, h = 12, level=95, 
                    pimethod = "nonparametric",
                    FUN=ahead::ridge2f))     
plot(fcast)                                 

(fcast <- forecast(object = fit_obj, h = 12, level=95, 
                    pimethod = "nonparametric",
                    FUN=ahead::varf))     
plot(fcast)