How Does Forecast Studio calculate forecast values for a Multiple Regression Model

[ridiculas]

Hi Guys

I set up a custom model (multiple linear regression model) in SAS Forecast Studio to get some coefficients for a few independent variables. Just trying to understand how SAS compute the forecast values becuase I did not put in any leading values for any independent variables. What is the equation used for this model?

Below is the code I copied from studio for the model of created, and also the prameter estimates in the attached file.

Thanks in advance.

Proc HPFARIMASPEC
/* Generated at: May 24, 2016 1:57:37 PM
* Model: MULTIREG
* Label: Y = CONST + Heating_Degree_Days ... + JUNE2012
*/
MODELREPOSITORY = work.temp
SPECNAME=MULTIPLEREGRESSION1
SPECLABEL=""
SPECTYPE=MULTIREG
SPECSOURCE=FSUI
;
FORECAST SYMBOL = Y TRANSFORM = NONE ;
INPUT SYMBOL = Heating_Degree_Days
TRANSFORM = NONE ;
INPUT SYMBOL = Cooling_Degree_Days
TRANSFORM = NONE ;
INPUT SYMBOL = PV_Cap_Log
TRANSFORM = NONE ;
INPUT SYMBOL = Max_of_Max_Temp
TRANSFORM = NONE ;
INPUT SYMBOL = SEASON
TRANSFORM = NONE ;
INPUT SYMBOL = Average_of_Avg_Humid
TRANSFORM = NONE ;
INPUT EVENT = JUNE2011
TRANSFORM = NONE ;
INPUT EVENT = JUNE2012
TRANSFORM = NONE ;
ESTIMATE
METHOD=CLS
CONVERGE=0.001
MAXITER=50
DELTA=0.001
SINGULAR=1.0E-7 ;
run;

[udo_sas · Accepted solution]

Hello -

If future values of independent variables are not included in the input data set, then the system will forecast those independent values using ESM. See STOCHASTIC statement of the HPFENGINE procedure for more information. SAS Forecast Studio will use future values of independent variables if they are included in the input data set.

Thanks,

Udo

[ridiculas]

Thanks udo. That answered by question.

I will check STOCHASTIC to find more info.