The COUNTREG Procedure

SPATIALDISPEFFECTS Statement

(Experimental)

  • SPATIALDISPEFFECTS <dispersion-spatial-effect-regressors> </options>;

The SPATIALDISPEFFECTS statement adds the spatially weighted dispersion-spatial-effect-regressors to regressors that are specified in the DISPMODEL statement. For example, if you specify q variables $z_{1}, \ldots , z_{q}$ in the SPATIALDISPEFFECTS statement, then each of q spatially weighted variables $\mathbf{W}z_{1}, \ldots , \mathbf{W}z_{q}$ has a parameter to be estimated in the regression. Here, $\mathbf{W}z_{1}, \ldots , \mathbf{W}z_{q}$ denotes the matrix and vector product between $\mathbf{W}$ and a column vector whose entries are the values of $z_{1}, \ldots , z_{q}$, respectively. The spatial weights matrix $\mathbf{W}$ comes from the data set that is specified in the WMAT= option in the PROC COUNTREG statement.

The "Parameter Estimates" table in the displayed output shows the estimates for spatially weighted explanatory variables; they are labeled with the prefix "Dsp_W_". For example, if you specify z (a variable in your data set) as a spatial effect explanatory variable, then the “Parameter Estimates” table labels the corresponding parameter estimate "Dsp_W_z".

You can specify the following option after a slash (/):

SELECT=INFO=<(selection-options)>
SELECTVAR=INFO=<(selection-options)>

requests that the variable selection method be based on an information criterion. For a list of selection-options, see the section Options for Variable Selection Based on an Information Criterion. For more information about this type of variable selection, see the section Variable Selection Using an Information Criterion.