SAS stores character data as one character per byte. You can use the LENGTH statement to increase or reduce the length of a character variable. You can also use other coding techniques to reduce the space that is needed for storing your character data.

SAS stores numeric data in floating-point representation. The default length for a numeric variable is 8 bytes. You can use a LENGTH statement to reduce the length of a numeric variable. Reading reduced-length numeric variables requires less I/O but more CPU resources than reading full-length numeric variables. You can use PROC COMPARE to see the precision loss, if any, in the values of reduced-length numeric variables.

By default, a SAS data file is uncompressed. You can compress your data files in order to conserve disk space, although some files are not good candidates for compression. Use the COMPRESS= data set option or system option to compress a data file. Use the POINTOBS=YES data set option to enable SAS to access observations in compressed files directly rather than sequentially. Use the REUSE=YES data set option or system option to specify that SAS should reuse space in a compressed file when observations are added or updated.

You can leave your data in its original storage location and use SAS data views to access the data in order to reduce the amount of space needed for storing data on disk. A DATA step view is a specific type of data view that is created in a DATA step with the VIEW= option. Use the DESCRIBE statement to write the source code for a data view to the SAS log. Some of the advantages of using DATA step views rather than data files are that they always access the most recent data in dynamic files and that they require less disk space. However, there can be an effect on performance when you use a DATA step view.