Plot the fit of an ARIMA Model

Hello,

I've created an ARIMA Model with the Econometrics Toolbox. I also easily managed to recreate this model manually with the arima and estimate functions. Now I want to predict the values using the model, plot it against the real values and calculate the RMSE of the predicted values. The econometric modeler does plot the fit of the model (seen in the pdf at Figure 1.1.) but I do not find any function to recreate this plot or the predicted values manually.

Is there an easy way to do this? Or do I have to manually include the equation myself?

Thank you in advance and kind regards

A. Enders

Answer 1

[Matthias Bär]

After digging 2 hours through different Matlab examples i finally found the answer. See below and try it.

load(fullfile(matlabroot,'examples','econ','Data_Airline.mat'))
y = log(Data);
T = length(y);

Mdl = arima('Constant',0,'D',1,'Seasonality',12,...
              'MALags',1,'SMALags',12);
EstMdl = estimate(Mdl,y);

residuals = infer(EstMdl,y);
prediction = y+residuals;

figure()
plot(y)
hold on
plot(prediction)

Good luck ;)

Cite: https://de.mathworks.com/matlabcentral/fileexchange/68576-stock-prediction-using-arima

[geofo]

This answer is not satisfactory. Let's say that the model is infering a vector of zero residuals. That means the prediction will be calculated as:

prediction = y

But this is not a prediction since the original data y is used to do the prediction.

[Jerry Yang]

residual = y - y_hat

so that y_hat = y - residual?

[Seemant Tiwari]

can you tell me, how are you calculating these value like constant (?), variance (?), sar (?), Sma (?). ??