Time series forecasting. Error with ARIMA model.

Hello. I have really strange situation. I have such simple code

    train_set = [1569;1062;2294;3239;1852;1274;3423;3623;1596]
    Mdl = arima(2,0,2);
    EstMdl = estimate(Mdl,train_set,'Display','off');

and I get error

Error using arima/validateModel (line 1314)
The non-seasonal moving average polynomial is non-invertible.

Error in arima/setLagOp (line 391)
Mdl = validateModel(Mdl);

Error in arima/estimate (line 1183)
Mdl = setLagOp(Mdl, 'MA' , LagOp([1  coefficients(iMA)' ], 'Lags', [0 LagsMA
]));

I thought that this is becase train_set vector is too short... but for train_set = [1;2;3;4;5;6;7;8;9] or train_set = sin(1:9).' or many others vectors everything works perfect. If I multiplay this vector by train_set = [train_set ; train_set ] it works fine too. Do anybody knows how it is possible ? How to ommit this error ?

[Brendan Hamm]

I am not able to reproduce your error:

train_set = [1569;1062;2294;3239;1852;1274;3423;3623;1596];
Mdl = arima(2,0,2);
EstMdl = estimate(Mdl,train_set)
Warning: Nonlinear inequality constraints are active; standard errors may be
inaccurate. 
> In arima/estimate (line 1113) 

      ARIMA(2,0,2) Model:
      --------------------
      Conditional Probability Distribution: Gaussian

                                    Standard          t     
       Parameter       Value          Error       Statistic 
      -----------   -----------   ------------   -----------
       Constant        3756.93       1674.45        2.24368
          AR{1}       0.156784      0.941963       0.166444
          AR{2}       -0.85079      0.320351       -2.65581
          MA{1}       0.747326      0.937987       0.796734
          MA{2}      -0.252674       1.43229      -0.176413
       Variance         124140        209476       0.592619
  EstMdl = 
      ARIMA(2,0,2) Model:
      --------------------
      Distribution: Name = 'Gaussian'
                 P: 2
                 D: 0
                 Q: 2
          Constant: 3756.93
                AR: {0.156784 -0.85079} at Lags [1 2]
               SAR: {}
                MA: {0.747326 -0.252674} at Lags [1 2]
               SMA: {}
          Variance: 124140