Assessing Efficiency Needs at Your Site

The first step in making an effective decision about how to optimize your SAS programs is to assess your site's technical environment, your program or programs, and your data.

Assessing Your Technical Environment

To determine which resources are scarce or costly at your site, work with your IT department to analyze the following characteristics of your technical environment:
Category
Characteristics
hardware
  • amount of available memory
  • number of CPUs
  • number and type of peripheral devices
  • communications hardware
  • network bandwidth
  • storage capacity
  • I/O bandwidth
  • capacity to upgrade
operating environment
  • resource allocation
  • scheduling algorithms
  • I/O methods
system load
  • number of users or jobs sharing system resources
  • network traffic (expected)
  • predicted increase in load in the future
SAS environment
  • which SAS software products are installed
  • number of CPUs and amount of memory that is allocated for SAS programming
  • which methods are available for running SAS programs at your site
In most cases, one or two resources are the most limited or most expensive for your programs. You can usually decrease the amount of critical resources that are used if you are willing to sacrifice some efficiency of the resources that are less critical at your site.

Assessing Your Programs

Developing an efficient program requires time and thought. To determine whether the additional amount of resources that are saved is worth the time and effort spent to achieve the savings, consider the following characteristics of each of your programs:
Characteristics
Guidelines for Optimizing
size of the program
As the program increases in size, the potential for savings increases. Focus on improving the efficiency of large programs.
number of times the program runs
The difference in resources used by an inefficient program and an efficient program that is run once or a few times is relatively small, whereas the cumulative difference for a program that is run frequently is large. Focus on improving the efficiency of programs that are run many times.

Assessing Your Data

The effectiveness of any efficiency technique depends greatly on your data. When you know the characteristics of your data, you can select the techniques that take advantage of those characteristics. Consider the following characteristics of your data:
Characteristic
Guidelines for Optimizing
volume of data
As the volume of data increases, the potential for savings also increases. Focus on improving the efficiency of programs that use large data sets or many data sets.
type of data
Specific efficiency techniques might work better with some types of data (for example, data that has missing values) than with others.
..................Content has been hidden....................

You can't read the all page of ebook, please click here login for view all page.
Reset
3.139.86.18