I am trying to do rolling windows volatility forecasting in gretl. In my dataset; 

I have 2409 obs. I will have an HAR model (rv5_SPX=c+rv5_SPX(-1)+harweek+harmonth] with 5 minutes realised volatility data and apply rolling windows forecast.

 I will estimate the HAR model with an initial sample 23-500 obs. and I am going to forecast 501. obs. Next, I will have the initial sample 24-501 obs. and forecast 502. obs. And again initial sample 25-502 and forecast 503. obs. ... up until 2000. obs. Each time, I need to reestimate my HAR model with initial sample and forecast next one, which is rolling windows forecasting. 

I could not find the codes for this purpose. Is it possible to do that in gretl?