forecast from a Markvov Switching regression model
res=ms_forecast(rms,hprev,exo_com,exo_idio)
* rms = a tlist result from a ms switching estimation
* hprev = the prevision period which can be either:
- a [n1 n2] constant vector where n1 and n2 are the lead over the last period of the estimation (n1<=0 means that the forecast begins within the estimation period)
- a n constant which is equivalent to [1 n] (forecast begins just after the estimation period)
- a [n1 n2] string vector where n1 and n2 are the time periods for forecasting (a posibility open only if the MS regression has been performed with ts)
- a n string which is equivalent to [1 n] (forecast begins just after the estimation period)
* exo_com = the data on the forecast horizon for the non switching exogenous variables that can be:
- a (H x n_x) matrix with H the number of forecasts and n_x the number of non switching exogenous variables
- a list of ts available over the forecasting horizon
- a list of (H x 1) vectors and ts available over the forecasting horizon
- a (H x n_x) string matrix of names (note: when the ms tlist results comes from a ms-mean or a var estimation, the variables are useless and can therefore be omitted)
* exo_idio = the data on the forecast horizon for the switching exogenous variables that can be:
- a (H x n_z) matrix with H the number of forecasts and n_z the number of switching exogenous variables
- a list of ts available over the forecasting horizon
- a list of (H x 1) vectors and ts available over the forecasting horizon
- a (H x n_x) string matrix of names (note: when the ms tlist results comes from a ms-mean or a var estimation, the variables are useless and can therefore be omitted)
* res = a results tlist with:
- res('meth') = 'msf'
- res('r_ms') = the tlist results from the MS estimation
- res('pstates') = the (H x nb_states) matrix of the probabilities of the respective states over the forecasting horizon
- res('prev_states') = the matrix of forecast of the endogenous variables accross the states
- res('prev') = the matrix of forecast of the endogenous variables
- res('hprev') = the forecast period