Skip to content

VARMA in R

4 messages · creepa1982, Giovanni Petris, Paul Gilbert

creepa1982
# 
Hi all, 

does anyone know of a package/function for fitting Vector Autoregressive
Moving Average models? I looked through most of the packages available but
could only find functions to fit a VAR. 

Any help would be appreciated! 

Benjamin
Giovanni Petris
# 
You may want to check package dlm and, possibly, dse. 

In dlm you can cast a VARMA model in state space form (dlmModARMA) and
estimate unknown parameters by maximum likelihood (dlmMLE). 


Best,
Giovanni

  
    

      
      
    

  


  


      

    

    

      
      

        


  

        

      Paul Gilbert
    

    

      
      #
    

  


  

      
Giovanni Petris wrote:

            

      
      
      
    

        
Yes, you can also do this in dse, either in the ARMA specification or as 
an equivalent state-space model. There is an example in the Users' Guide 
distributed with the package.

Paul

            

      
      
      
    

        

      
      
    

            

      
      
      
    

        
====================================================================================

La version fran?aise suit le texte anglais.

------------------------------------------------------------------------------------

This email may contain privileged and/or confidential in...{{dropped:26}}

  
  


  


      

    

    

      
      

        


  

        

      creepa1982
    

    

      
      #
    

  


  

      
Hey Giovanni, 

thanks a lot for the help. I tried out combining the two functions
dlmModARMA and dlmMLE and it works. The only problem I have right now is
this. When I pass on the information about the starting parameters (param)
in the dlmMLE function I can only input one parameter vector. However, for a
VARMA I have a matrix of coefficients for both the AR part and MA part. How
can I signal to dlmModARMA which part of the passed on vector is supposed to
be the AR input, the MA input and so on? 

Thanks again! 

Benjamin

        
Giovanni Petris wrote: