gretl version 2017d-git
Current session: 2017-09-07 07:56

? open phillips_aus.gdt # poe-4

Read datafile C:\Users\Lenovo\AppData\Roaming\gretl\data\POE4\phillips_aus.gdt
periodicity: 4, maxobs: 91
observations range: 1987:1 to 2009:3

Listing 3 variables:
  0) const    1) inf      2) u      

? function void trytry(series y, list xlist[null])
> arma 1 1; y xlist
> arma 1 1; y xlist
> end function
? list x = 0 diff(u)
Generated list x
? trytry(inf,x)

Function evaluations: 132
Evaluations of gradient: 22

ARMAX, using observations 1987:2-2009:3 (T = 90)
Estimated using Kalman filter (exact ML)
Dependent variable: y
Standard errors based on Hessian

             coefficient   std. error     z       p-value 
  --------------------------------------------------------
  const        0.715975    0.233031      3.072   0.0021    ***
  phi_1        0.903959    0.0674654    13.40    6.14e-041 ***
  theta_1     −0.548921    0.123771     −4.435   9.21e-06  ***
  d_u         −0.820599    0.231340     −3.547   0.0004    ***

Mean dependent var   0.791111   S.D. dependent var   0.636819
Mean of innovations  0.010436   S.D. of innovations  0.484094
Log-likelihood      −62.75480   Akaike criterion     135.5096
Schwarz criterion    148.0086   Hannan-Quinn         140.5499

                        Real  Imaginary    Modulus  Frequency
  -----------------------------------------------------------
  AR
    Root  1           1.1062     0.0000     1.1062     0.0000
  MA
    Root  1           1.8218     0.0000     1.8218     0.0000
  -----------------------------------------------------------


Function evaluations: 108
Evaluations of gradient: 22

ARMAX, using observations 1987:2-2009:3 (T = 90)
Estimated using Kalman filter (exact ML)
Dependent variable: y
Standard errors based on Hessian

             coefficient   std. error     z       p-value 
  --------------------------------------------------------
  const        0.872373    0.224883      3.879   0.0001    ***
  phi_1        0.912143    0.0610349    14.94    1.69e-050 ***
  theta_1     −0.600730    0.112533     −5.338   9.39e-08  ***
  d_u         −0.772800    0.226656     −3.410   0.0007    ***

Mean dependent var   0.791111   S.D. dependent var   0.636819
Mean of innovations −0.026602   S.D. of innovations  0.483563
Log-likelihood      −62.63453   Akaike criterion     135.2691
Schwarz criterion    147.7681   Hannan-Quinn         140.3094

                        Real  Imaginary    Modulus  Frequency
  -----------------------------------------------------------
  AR
    Root  1           1.0963     0.0000     1.0963     0.0000
  MA
    Root  1           1.6646     0.0000     1.6646     0.0000
  -----------------------------------------------------------

? include lagreg.gfn
 C:\Users\Lenovo\AppData\Roaming\gretl\functions\lagreg\lagreg.gfn
lagreg 0.61, 2017-02-13 (Oleh Komashko (code by Allin Cottrell used))
? auto_arma(inf,x)

Best aic model: SARMAX(1,1;0,0)

Test for autocorrelation up to order 4

Ljung-Box Q' = 0.727663,
with p-value = P(Chi-square(2) > 0.727663) = 0.6950


Best bic model: SARMAX(1,1;0,0)

Test for autocorrelation up to order 4

Ljung-Box Q' = 0.733717,
with p-value = P(Chi-square(2) > 0.733717) = 0.6929


Best hqc model: SARMAX(1,1;0,0)

Best aicc model: SARMAX(1,1;0,0)

? auto_arma(inf,x,2,0)

Best aic model: ARMAX(1,1)

Test for autocorrelation up to order 4

Ljung-Box Q' = 0.733572,
with p-value = P(Chi-square(2) > 0.733572) = 0.6930


Best bic model: ARMAX(1,1)

Best hqc model: ARMAX(1,1)

Best hqc model: ARMAX(1,1)

# 2017b-git 2017-04-24