SAS Companion for the OS/390 Environment

Estimating the Size of a SAS Data Set

Under OS/390, a non-HFS SAS data library is a single file that contains the library's SAS data sets. You cannot obtain information from the file system on how large each member SAS data set is.

To obtain a rough estimate of how much space you need for a disk-format, non-HFS SAS data set that was created by the V8 engine, follow these steps:

Note: This procedure is valid only for uncompressed native SAS library members that were created with the V8 engine. [cautionend]

Using the CONTENTS procedure, specify the DIRECTORY option and using the DATA option to specify the SAS library member.
Using the output of the CONTENTS procedure, divide the Data Set Page Size statistic for the SAS library member by the Blocksize statistic for the directory to obtain the number of blocks per page.
Multiply the number of blocks per page by the Number of Data Set Pages statistic for the SAS library member to obtain the number of blocks.
Increase the number of blocks by 5% to account for overhead associated with the library directory control structures associated with the SAS library member.
Divide the increased number of blocks by the Blocks Per Track statistic for the directory, then round up to the nearest whole track.

Note:

The number of tracks required for a member of a SAS data library can vary depending on the track size of the device on which the data library resides, on the block size of the SAS data library's operating environment data set, and on the data set page size for the member.
The Max Obs per Page statistic for the member may be used to adjust the Number of Data Set Pages statistic in step 3 above to estimate the space required to add a number of new observations. Divide the number of new observations by the Max Obs per Page statistic to get the number of new data set pages needed. Then add the number of new pages to the Number of Data Set Pages statistic.

To determine the rough size in bytes of the allocation needed for a SAS library member, follow these steps:

Run the CONTENTS procedure using the DATA option to specify the SAS library member.
Using the output of the CONTENTS procedure, multiply the Pagesize value for the SAS library member by the number of pages for the member to obtain a number of bytes.
Increase the number of bytes by 5% to account for overhead associated with the library directory control structures associated with the library member.

For additional information about estimating the size of a SAS data set, see Tuning SAS Applications in the MVS Environment, by Michael Raithel. This book is available from SAS Institute as part of the Books by Users program.

Using the CONTENTS Procedure to Determine Observation Length

To determine the length of each observation in a Version 8 SAS data set, you can create a Version 8 SAS data set that contains one observation. Then run the CONTENTS procedure to determine the observation length. The CONTENTS procedure displays engine/host-dependent information, including page size and the number of observations per page for uncompressed SAS data sets. For example, the following input produces a SAS data set plus PROC CONTENTS output:

data oranges;
input variety $ flavor texture looks;
cards;
navel 9 8 6 ;
proc contents data=oranges; 
run;

The output is shown in CONTENTS Procedure Output.

CONTENTS Procedure Output

The SAS System 1 The CONTENTS Procedure Data Set Name: WORK.ORANGES Observations: 1 Member Type: DATA Variables: 4 Engine: V8 Indexes: 0 Created: 14:27 Tuesday, March 5, 1999 Observation Length: 32 Last Modified: 14:27 Tuesday, March 5, 1999 Deleted Observations: 0 Protection: Compressed: NO Data Set Type: Sorted: NO Label: -----Engine/Host Dependent Information----- Data Set Page Size: 6144 Number of Data Set Pages: 1 First Data Page: 1 Max Obs per Page: 139 Obs in First Data Page: 4 Number of Data Set Repairs: 0 Physical Name: SYS96065.T142625.RA000.USERID.R0000180 Release Created: 7.0000B2 Release Last Modified: 7.0000B2 Created by: USERID Last Modified by: USERID Subextents: 1 Total Blocks Used: 1 -----Alphabetic List of Variables and Attributes----- # Variable Type Len Pos Format Informat Label ----------------------------------------------------------------------------------- 2 FLAVOR Num 8 8 4 LOOKS Num 8 24 3 TEXTURE Num 8 16 1 VARIETY Char 8 0

The only values that you need to pay attention to are Observation Length and Compressed:

Observation Length: is the record size in bytes.
Compressed: has the value NO if records are not compressed; it has the value YES if records are compressed. (If the records are compressed, do not use the procedure given in Estimating the Size of a SAS Data Set.)

Chapter Contents
Previous
Next
Top of Page