assign estimated values

Previous Topic Next Topic
 
classic Classic list List threaded Threaded
4 messages Options
Reply | Threaded
Open this post in threaded view
|

assign estimated values

malaka
This post has NOT been accepted by the mailing list yet.
CONTENTS DELETED
The author has deleted this message.
Reply | Threaded
Open this post in threaded view
|

Re: assign estimated values

Pete Brecknock
malaka wrote
Hi,
I want to assign the ar1 , ma 1 and the intercept estimated by the following code to three variables a, b and c respectively.

Can anyone help me with this please?

code:

a0 = 0.05; a1 = 0.1; b1 = 0.85
nu = rnorm(2500)
epsi = rep(0, 2500)
h = rep(0, 2500)
for (i in 2: 2500) { h[i] = a0 + a1 * epsi[i-1]^2 + b1 * h[i-1]  ; epsi[i] = nu[i] * sqrt(h[i])}
epsi = epsi[1501:2500]
epsi=epsi*epsi
arma(epsi,order=c(1,1))
You haven't specified a library for the arma function. Your code doesn't work for me in the form you posted it.

However, changing ...

arma(epsi, order=c(1,1)) to

mod = arima(epsi, order=c(1,0,1))

You can extract the parameters of interest using

a= mod$coef["ar1"]
b= mod$coef["ma1"]
c= mod$coef["intercept"]

HTH

Pete
Reply | Threaded
Open this post in threaded view
|

Re: assign estimated values

malaka
This post has NOT been accepted by the mailing list yet.
CONTENTS DELETED
The author has deleted this message.
Reply | Threaded
Open this post in threaded view
|

Re: assign estimated values

Rui Barradas
In reply to this post by Pete Brecknock
Hello,

Em 11-02-2013 01:26, Pete Brecknock escreveu:

> malaka wrote
>> Hi,
>> I want to assign the ar1 , ma 1 and the intercept estimated by the
>> following code to three variables a, b and c respectively.
>>
>> Can anyone help me with this please?
>>
>> code:
>>
>> a0 = 0.05; a1 = 0.1; b1 = 0.85
>> nu = rnorm(2500)
>> epsi = rep(0, 2500)
>> h = rep(0, 2500)
>> for (i in 2: 2500) { h[i] = a0 + a1 * epsi[i-1]^2 + b1 * h[i-1]  ; epsi[i]
>> = nu[i] * sqrt(h[i])}
>> epsi = epsi[1501:2500]
>> epsi=epsi*epsi
>> arma(epsi,order=c(1,1))
>
> You haven't specified a library for the arma function. Your code doesn't
> work for me in the form you posted it.

If it's function arma(9 in package tseries, then your solution below
also works. tseries::arma returns an object of class "arma",

mod <- arma(epsi, order=c(1,1))
str(mod)
List of 12
  $ coef             : Named num [1:3] 0.971 -0.9214 0.0325
   ..- attr(*, "names")= chr [1:3] "ar1" "ma1" "intercept"
  $ css              : num 2416
  $ n.used           : int 1000
  $ residuals        : num [1:1000] NA 0.105 0.1747 0.0144 0.0435 ...
  $ fitted.values    : Time-Series [1:1000] from 1 to 1000: NA 0.175
0.208 0.243 0.269 ...
  $ series           : chr "epsi"
  $ frequency        : num 1
  $ call             : language arma(x = epsi, order = c(1, 1))
  $ vcov             : num [1:3, 1:3] 0.000203 -0.00028 -0.000224
-0.00028 0.00053 ...
   ..- attr(*, "dimnames")=List of 2
   .. ..$ : chr [1:3] "ar1" "ma1" "intercept"
   .. ..$ : chr [1:3] "ar1" "ma1" "intercept"
  $ lag              :List of 2
   ..$ ar: int 1
   ..$ ma: int 1
  $ convergence      : int 0
  $ include.intercept: logi TRUE
  - attr(*, "class")= chr "arma"


mod$coef is a named vector, with names "ar1", "ma1" and "intercept".


Hope this helps,

Rui Barradas

>
> However, changing ...
>
> arma(epsi, order=c(1,1)) to
>
> mod = arima(epsi, order=c(1,0,1))
>
> You can extract the parameters of interest using
>
> a= mod$coef["ar1"]
> b= mod$coef["ma1"]
> c= mod$coef["intercept"]
>
> HTH
>
> Pete
>
>
>
> --
> View this message in context: http://r.789695.n4.nabble.com/assign-estimated-values-tp4658139p4658141.html
> Sent from the R help mailing list archive at Nabble.com.
>
> ______________________________________________
> [hidden email] mailing list
> 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.
>

______________________________________________
[hidden email] mailing list
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.