The non-seasonal moving average polynomial is non-invertible

Hi everyone,

I am trying to simulate and estimate an arma(3,3) process:

model = arima('Constant',0.5,'AR',{0.5;0.3;0.01},'MA',{0.3;0.2;0.1},'Variance',1);
rng('shuffle')
Y = simulate(model,1000);
mod=arima(3,0,3);
[EstMdl] = estimate(mod,Y);

I've created a loop to do this 1000 times. After some iterations (I've already got 300 results before the error message), the following error message appears:

Error using arima/validateModel (line 1290)
The non-seasonal autoregressive polynomial is unstable.
Error in arima/setLagOp (line 401)
   OBJ = validateModel(OBJ);
Error in arima/estimate (line 1086)
  OBJ = setLagOp(OBJ, 'AR' , LagOp([1 -coefficients(iAR)' ], 'Lags', [0 LagsAR ])

Why does this happen? And why this happen only after some iterations? What can I do to avoid the error?

Best regards

Roberto

[Seemant Tiwari]

hi,

can you tell me, how are you calculating these values?

'constant', o.5, 'ar',{ 0.5,0.3,0.1}, 'ma', {0.3,,0.2,0.1} ??

i have 1 year hourlywind speed data 365x24 = 8760

i have calculate p, d, q value my p value is 1, q value is 0 and d value is 1.

now i want to create model but i am not understanding how can i calculate 'AR' MA' values.

Thank you

Answer 1

[Hang Qian]

Hi Roberto,

The error message “the non-seasonal autoregressive polynomial is unstable” indicates that some of the eigenvalues of the AR part of the series are outside the unit circle, hence non-stationarity. Similarly, the error “the non-seasonal moving average polynomial is non-invertible” shows explosion of the MA roots, hence non-invertibility.

Practically it is not a good idea to fit a high-order ARMA model. It might require a huge sample size to provide a good estimate on a ARMA(3,3) model, even if the data are simulated from the true model. To avoid those error messages, be parsimonious in model specification, say ARMA(1,1). If you really want a high-order model, you may consider informative priors to shrink the coefficients.

Thank you.

Regards,

Hang Qian

[Seemant Tiwari]

hello,

i have hourly wind speed data, no. of data is 8760

i have calculate p, d, q value

p = 1

d = 0

q = 1

now i want to create arima model.

can you tell me, how can i calculate these value

MD =ARIMA(CONSTANT (? ), 'AR (?), 'SAR' (?), 'SMA' (?), 'VARIANCE' (?))

Thank you