Skip to content

extracting coefficients from ar() output

7 messages · T.Riedle, William Dunlap, Peter Dalgaard +1 more

T.Riedle
# 
Hi everybody,

I am trying to run an AR1 model using the ar() function as shown below.
order ar        var.pred x.mean   aic        n.used order.max partialacf resid      method        series
 [1,] 1     0.7433347 1.382908 49.99861 Numeric,16 36     15        Numeric,15 Numeric,36 "Yule-Walker" "data"
 [2,] 1     0.7410181 1.565755 49.94778 Numeric,16 36     15        Numeric,15 Numeric,36 "Yule-Walker" "data"
 [3,] 1     0.7636966 1.660581 49.86861 Numeric,16 36     15        Numeric,15 Numeric,36 "Yule-Walker" "data"


I get the table as shown above if I use head().

How can I extract the ar coefficients from this table? I have already tried coef() and rollingarma$ar but both do not work.
What can I do?

Thanks for your help.
William Dunlap
# 
help(ar) should tell you how to get the coefficients.  If, like me, you
don't
read help files, you can use str() to look at the structure of ar's output.
List of 14
 $ order       : int 2
 $ ar          : num [1:2] 1.011 -0.918
 $ var.pred    : num 0.0654
 $ x.mean      : num 0.00934
 $ aic         : Named num [1:15] 61.215 53.442 0 0.985 2.917 ...
  ..- attr(*, "names")= chr [1:15] "0" "1" "2" "3" ...
 $ n.used      : int 30
 $ order.max   : num 14
 $ partialacf  : num [1:14, 1, 1] 0.5273 -0.9179 -0.1824 -0.0477 -0.0393 ...
 $ resid       : num [1:30] NA NA -0.0145 -0.0734 -0.0725 ...
 $ method      : chr "Yule-Walker"
 $ series      : chr "sin(1:30)"
 $ frequency   : num 1
 $ call        : language ar(x = sin(1:30), aic = TRUE)
 $ asy.var.coef: num [1:2, 1:2] 0.00583 -0.00308 -0.00308 0.00583
 - attr(*, "class")= chr "ar"
[1]  1.0112512 -0.9178554




Bill Dunlap
TIBCO Software
wdunlap tibco.com
On Thu, Jun 16, 2016 at 4:34 AM, T.Riedle <tr206 at kent.ac.uk> wrote:

            

      
      
      
    

        

      
      
    

        

  
  


  


      

    

    

      
      

        


  

        

      Peter Dalgaard
    

    

      
      #
    

  


  

          

      
      
      
    

        
Also notice that the output of rollapply is not an ar object. More likely a list of them, so  try rollingarma[[i]]$ar or maybe lapply(rollingarma, function(x)x$ar) or sapply(rollingarma, "[[", "ar") or...

            

      
      
      
    

        

      
      
    

            

      
      
      
    

        

      
      
    

        

  
    

      
      
    

  


  


      

    

    
1 day later

    

      
      

        


  

        

      T.Riedle
    

    

      
      #
    

  


  

      
Thank you very much, Peter. I played a bit and found a solution.

            

      
      
      
    

        
[1] 0.9430692 0.9140253 0.9236898 0.9426744 0.9465110 0.9318470 0.9033054 0.9206048 0.9243736 0.9129082
[11] 0.9181811 0.9350779 0.9464205 0.9410245 0.9335568 0.9201928 0.8869414 0.8320984 0.8185671 0.7989182
[21] 0.7454876 0.6388364 0.6797046 0.6704642 0.7077033 0.8895698 0.8755445 0.8965050 0.8969068 0.8891385
[31] 0.9284835 0.9628297 0.9674624 0.9524462 0.9423693 0.9629843 0.9996613 1.0000295 0.9845222 0.9877242
[41] 0.9582863 0.9596756 0.9415847 0.9471677 0.9447052 0.9324048 0.9171082 0.8928825 0.9133751 0.9203662

I am trying to export the data to Excel using WriteXLS:

            

      
      
      
    

        
Unfortunately, it doesn't work. How can I export the data to Excel?

-----Original Message-----
From: peter dalgaard [mailto:pdalgd at gmail.com] 
Sent: 16 June 2016 18:49
To: William Dunlap
Cc: T.Riedle; R-help at r-project.org
Subject: Re: [R] extracting coefficients from ar() output

            

      
      
      
    

        
Also notice that the output of rollapply is not an ar object. More likely a list of them, so  try rollingarma[[i]]$ar or maybe lapply(rollingarma, function(x)x$ar) or sapply(rollingarma, "[[", "ar") or...

            

      
      
      
    

        

      
      
    

            

      
      
      
    

        

      
      
    

        
--
Peter Dalgaard, Professor,
Center for Statistics, Copenhagen Business School Solbjerg Plads 3, 2000 Frederiksberg, Denmark
Phone: (+45)38153501
Office: A 4.23
Email: pd.mes at cbs.dk  Priv: PDalgd at gmail.com

  
  


  


      

    

    

      
      

        


  

        

      David Winsemius
    

    

      
      #
    

  


  

          

      
      
      
    

        
Just a guess (since you didn't include the error message) but do you have a length-1 character vector named `test`. I thought not. So try using SheetNames = 'test'

  
    

      
      
    

  


  


      

    

    

      
      

        


  

        

      T.Riedle
    

    

      
      #
    

  


  

      
Thank you very much.
Here is the error message.

            

      
      
      
    

        
Error in WriteXLS(coefar, ExcelFileName = "R.xls", SheetNames = "test") : 
  'x' must be the name of a data frame, the name of a list of data frames, a data frame object, a list object of data frames.

-----Original Message-----
From: David Winsemius [mailto:dwinsemius at comcast.net] 
Sent: 17 June 2016 19:27
To: T.Riedle
Cc: peter dalgaard; R-help at r-project.org
Subject: Re: [R] extracting coefficients from ar() output

            

      
      
      
    

        
Just a guess (since you didn't include the error message) but do you have a length-1 character vector named `test`. I thought not. So try using SheetNames = 'test'

--
David.

            

      
      
      
    

        

      
      
    

            

      
      
      
    

        

      
      
    

            

      
      
      
    

        

      
      
    

        
David Winsemius
Alameda, CA, USA

  
  


  


      

    

    

      
      

        


  

        

      David Winsemius
    

    

      
      #
    

  


  

          

      
      
      
    

        
sos::findFn("writeXLS")

So after tracking down the package that has writeXLS (also named writeXLS), I see that the example in the help page quotes the names of dataframe objects and the help page says the x values should be _character_ (rather than R symbol/names). You didn't provide a character vector as the first argument which, since it was unnamed, was therefore matched to the 'x' parameter.