Forecast evaluation

[j52]

Hello,

is there an option to generate a forecast evaluation of two imported series, e.g. to compare the actual generated forecast (w/o using eviews) of a company to the actual sales?

Thanks a lot.

[EViews Gareth]

Which forecast evaluation would you like to perform?

[j52]

Root Mean Square Error, Mean Absolute Error, Mean Abs. Percent Error, Theil Inequality Coefficient.

[EViews Gareth]

Code: Select all

@rmse(y,yf) @mae(y,yf) @mape(y,yf) @theil(y,yf)

[j52]

Awesome, thanks a lot!

[j52]

Might be a stupid question, buT how do I show the value? Do I have to create a series: series RMSE=@RMSE(d,df)? Or is there a faster way to show the one or even all of the evaluations at once? And what the command for the Theil's Bias, Variance and Covariance Proportion? Thanks a lot.

[EViews Gareth]

Just type:

Code: Select all

=@rmse(y,yf)

There is no command for the components.

[Nicole]

Hi,

I am using Eviews 8. When I try to calculate the RMSE and MAE for my forecasts using the @mae(return,returnf) and @rmse(return,returnf), but I keep getting the same number for both measurement errors. How can this be? Am I typing it incorrectly? :?

[EViews Gareth]

Code: Select all

create u 10 series return=nrnd series returnf=nrnd scalar rmse = @rmse(return, returnf) scalar mae = @mae(return, returnf) show rmse show mae

Run that a few times, you should get different numbers each time.