CSPATIALREG Procedure

The SPATIALID statement specifies a variable that identifies a spatial unit for each observation in the two data tables that you specify in the DATA= and WMAT= options in the PROC CSPATIALREG statement. The variable that you specify in the SPATIALID statement is also used to match the rows and columns in the spatial weights matrix. Only one SPATIALID statement and one spatial ID variable are allowed. For all models except a purely linear regression model without an SLX effect, you need a SPATIALID statement in order to specify a spatial ID variable for the purpose of matching observations in two data tables that you specify in the DATA= and WMAT= options. The values of the spatial ID variable in either the DATA= or WMAT= data table cannot be missing.

The variable in the SPATIALID statement can be either numeric or character. However, the type of spatial ID variable in the two data tables that you specify in the DATA= and WMAT= options must be the same. When the spatial ID variable is numeric, it needs to be integer-valued. If you specify a number that is not an integer, PROC CSPATIALREG uses the integer part of that number for matching.