FOCUS AREAS

Step 2

Migration

Best Practices Migration with PROC MIGRATE: Migrate Files

Special Topic: Migrating Linear Model Item Stores

The STORE statement in many SAS/STAT linear modeling procedures can save the context and results of a statistical analysis into an item store. Item stores might not be usable if you change to an incompatible operating environment, because item stores are not supported by cross-environment data access (CEDA). (See the CEDA topic in SAS Language Reference: Concepts.)

You cannot use PROC MIGRATE and Remote Library Services (RLS) to migrate a linear model item store to a different environment. However, if the procedure stored the SAS code that created the item store, you can use the PLM procedure in the source environment to show the code, and then submit that code in the target environment.

  1. In the source environment, run PROC PLM with SHOW PROGRAM to show the SAS code from the item store. In the example code below, the item store is named Cheese to match PROC PLM examples in the user's guide.

    The FILE argument in the ODS LISTING statement writes the output to a file (Mycode.lst here).

      libname Srclib "c:\mydata";
      ods listing file="c:\mydata\Mycode.lst";
      proc plm restore=Srclib.Cheese;
        show program;
      run;
      ods listing close;
    
  2. Open the listing output program (Mycode.lst) in a text editor. The output contains a section titled "SAS Program Information." You can copy that SAS code and paste it into a session to the target environment. Remember that you also need the original data. Because the SAS code contains the STORE statement, it re-creates the item store in the target environment.

The item stores produced by the following SAS/STAT procedures store the analysis program code. You can use the instructions described above.

GLMSELECT, LIFEREG, LOGISTIC, MIXED, ORTHOREG, SURVEYREG, SURVEYLOGISTIC

The item stores produced by the following SAS/STAT procedures do not store the analysis program code. You cannot use the instructions described above.

GENMOD, GLIMMIX, GLM, PHREG, PROBIT, SURVEYPHREG

For more information, see the STORE statement and PROC PLM in the SAS/STAT User's Guide.

Return to Step 2.