Dependent Variable: LOG(N)
Method: ARMA Maximum Likelihood (OPG - BHHH)
Date: 06/28/20 Time: 13:59
Sample: 2017M01 2019M12
Included observations: 36
Failure to improve objective (non-zero gradients) after 363 iterations
Coefficient covariance computed using outer product of gradients
Variable Coefficient Std. Error t-Statistic Prob.
AR(1) 0.996473 0.052603 18.94329 0.0000
SAR(12) 1.000000 2.39E-05 41843.38 0.0000
MA(1) -0.836555 0.272927 -3.065128 0.0045
SMA(12) -0.999685 0.000126 -7916.131 0.0000
SIGMASQ 0.126676 0.027721 4.569657 0.0001
R-squared 0.444712 Mean dependent var 6.404303
Adjusted R-squared 0.373061 S.D. dependent var 0.484401
S.E. of regression 0.383546 Akaike info criterion 1.686063
Sum squared resid 4.560333 Schwarz criterion 1.905997
Log likelihood -25.34914 Hannan-Quinn criter. 1.762826
Durbin-Watson stat 1.386845
Inverted AR Roots 1.00 1.00 .87+.50i .87-.50i
.50+.87i .50-.87i .00+1.00i -.00-1.00i
-.50+.87i -.50-.87i -.87-.50i -.87+.50i
-1.00
Estimated AR process is nonstationary
Inverted MA Roots 1.00 .87+.50i .87-.50i .84
.50+.87i .50-.87i .00+1.00i -.00-1.00i
-.50+.87i -.50-.87i -.87-.50i -.87+.50i
-1.00
我估计的模型命令是 log(n)ar(1)sar(12)ma(1)sma(12)不知道这个结果模型可否通过啊