-    
Hi, hope that you will be fine, I have a problem with functional time series,   
I am working with the hourly electricity spot price data, due to the large 
dimensionality I convert the discrete data into functional data. The model I 
use the functional autoregressive model of order more than one, as per my 
knowledge there no R package available to deal with such model, so I apply an 
alternative method, using the functional principal components (FPC's) as 
dimensional reduction, utilizing the associated principal components for the 
forecasting through multivariate time series model. Then I convert these 
forecast scores into functional curves through Karhunen-Loeve decomposition 
into a functional form, in such a way I obtained a forecast of each day as a 
single curve. Know, to check the accuracy of the model I want to calculate 
percentage mean square error or mean absolute error. know my problem start from 
here, So I want to reverse back each curve into 24 discrete points, is there is 
any package in R which is helpful in dealing with such a problem. I will be 
waiting
  for your fruitful reply in this regard.

   - 


        [[alternative HTML version deleted]]

______________________________________________
R-help@r-project.org mailing list -- To UNSUBSCRIBE and more, see
https://stat.ethz.ch/mailman/listinfo/r-help
PLEASE do read the posting guide http://www.R-project.org/posting-guide.html
and provide commented, minimal, self-contained, reproducible code.

Reply via email to