The LP Procedure

Below are statements used in PROC LP, listed in alphabetical order as they appear in the text that follows.

The TYPE, ID (or ROW), VAR, RHS, RHSSEN, and RANGE statements are used for identifying variables in the problem data set when the model is in the dense input format. In the dense input format, a model’s variables appear as variables in the problem data set. The TYPE, ID (or ROW), and RHS statements can be omitted if the input data set contains variables _TYPE_, _ID_ (or _ROW_), and _RHS_; otherwise, they must be used. The VAR statement is optional. When it is omitted, PROC LP treats all numeric variables that are not explicitly or implicitly included in RHS, RHSSEN, and RANGE statements as structural variables. The RHSSEN and RANGE statements are optional statements for sensitivity and range analyses. They can be omitted if the input data set contains the _RHSSEN_ and _RANGE_ variables.

The TYPE, COL, ROW (or ID), COEF, RHS, RHSSEN, and RANGE statements are used for identifying variables in the problem data set when the model is in the sparse input format. In the sparse input format, a model’s rows and columns appear as observations in the problem data set. The TYPE, COL, ROW (or ID), and COEF statements can be omitted if the input data set contains the _TYPE_ and _COL_ variables, as well as variables beginning with the prefixes _ROW (or _ID) and _COEF. Otherwise, they must be used. The RHS, RHSSEN, and RANGE statements identify the corresponding columns in the model. These statements can be omitted if there are observations that contain the RHS, RHSSEN, and RANGE types or the _RHS_, _RHSSEN_, and _RANGE_ column values.

The SHOW, RESET, PRINT, QUIT, PIVOT, IPIVOT, and RUN statements are especially useful when executing PROC LP interactively. However, they can also be used in batch mode.