unixprompt% spss -m filename.spx > output.out

Both SPSS and SAS syntax program require editing before they can be run on a user's computer system. In SPSS, the location of the raw data file, specified in the "Data List File" command line, must be indicated:

data list file = "filename.raw"

The UNIX path is typically indicated as:

data list file = "/your path/filename.raw"

The specific PC path may be indicated as:

data list file = "c:\my documents\filename.raw"

(Please note that your path and my documents are used for illustrative purposes only, the actual path name depends on where the customer has housed the raw data file.)
For SAS files, three lines of syntax must be edited before the batch program can be run in SAS. Again, the exact path indicating the location of the raw data file must be added to the FILENAME line. The LIBNAME statement appears at the top of the syntax file.

LIBNAME LIBRARY '/your path';

The DATA and FILENAME statements appear after the VALUES statements, just before the INPUT statement:

DATA LIBRARY.datasetname;

FILENAME DATAFILE 'filename.raw';

After editing the various lines of syntax code in the programs, they are ready to run in batch mode.

Customers should note that many of the SSEDL SPSS and SAS syntax programs are too large to run in batch mode on the Windows versions of SPSS and SAS. The size of a dataset often precludes processing the entire syntax program file in a single executed run because of excessive "syntactic elements." It may not be possible to process an entire syntax program without some editing to the program. We recommend the following:

For SPSS Windows:

1. Run pieces of the syntax at a time. For some datasets we recommend running the data list portion of the syntax first, adding an execute command on the next line after the period that delimits the DATA LIST section of the syntax. After running the Data List syntax (which reads in the raw data and names the variables), the active file should be saved. After saving the system file, run the remaining sections of syntax such as the VARIABLE LABELS or VALUE LABELS commands. This is accomplished by placing the EXECUTE command on the next line after the period that delimits the VARIABLE LABELS, VALUE LABELS, MISSING VALUES, or RECODES section of syntax. Sections of syntax may be executed by highlighting the section of syntax and then issuing the "run selection" command from the menu. If a large command, such as VALUE LABELS, is run in pieces, a new Value Labels command must be placed at the beginning of each block of labels that will be processed. It is recommended that the active system file be saved periodically. For extremely large syntax files, modifications may be required because the syntax program contains too many syntactic elements for SPSS/Windows to process in a single run.

If you attempt to run the entire syntax program, or even smaller pieces of the program, and receive the message "missing command delimiter," the problem may be that SPSS cannot process so many pieces of syntax in a single batch or executed run.

2. Create syntax extracts with a smaller selection of variables. This may require extensive editing.

For larger datasets, if you have access to a network or mainframe system with SPSSx, it is often easier to create the analysis file on that system, and then create a portable system file. SPSS Windows can import a large SPSSx portable system file. SPSS Windows doesn't have problems reading enormous portables files, but does give problems when syntax files have many, many commands and declarations.

For SAS:

Remove the individual FORMAT declarations toward the bottom of the syntax file. SAS/Windows does limit the number of unique FORMATS that can be expressed. Formats and label assignments may be accomplished in another program run.

Return to the Help/FAQ Page

or

Go to the Next Question