Support DSGE-VAR forecasts

Currently, forecasts are based on the DSGE model and not the DSGE-VAR

@stepan-a Can you confirm this? Also, it is nowhere documented, so this borders on a bug. I would like to better state what we are doing in the manual.

@JohannesPfeifer yes, we did not implement forecasts from the dsge-var model (only IRFs compared with those of the DSGE). We should add a remark in the documentation for the forecastcommand. We have another command for computing forecasts from BVAR models (bvar_forecast), we may add a new command dsgevar_forecast. This would not be too much pain because we already have codes for computing the hyper parameters of the Gaussian Inverse Wishart prior of the DSGE-VAR, but I do not think that this is a priority...

Created by: npak243

@stepan-a : Sorry for jump in this conversation but do you have any plan to develop dsgevar_forecast command? Thank for your information so much.

@npak243 I do not really have plans about this... I did not look into this part since years. I will see what I can do when I deal with #1021 (closed), but if you feel that this is important you are welcome to contribute code (if you do a pull request for the matlab code, we will add the interface in the preprocessor). Basically, for a point forecast (based on the posterior mode for instance), we should just have to call the dsgevar_posterior_density routine to obtain the Gaussian-Inverse Wishart posterior for the VAR model, and then simulate this BVAR starting from historical initial conditions.

Created by: mona-samy

could you please send me matlab code of dsge ,dsge- var,dsge- dfm , dfm- var models