Paper 137

RIGHT-SIZING CHARACTER VARIABLES: A SAS DATA MANAGEMENT STRATEGY FOR OPTIMIZING LARGE DATASETS

September 23, 2025: 1:30 PM - 2:00 PM

Data Collection, Management, and Manipulation (DCMM), Executive Dining Room #3

Authors	Abstract
Divya Gadde	Managing large, raw datasets in SAS—especially those lacking metadata or proper layout documentation—often leads to inefficient data structures, inflated storage requirements, and slow performance. This paper presents a practical and scalable solution using the %RESIZE macro, designed to optimize character variable lengths by dynamically adjusting them to their actual maximum content length. The %RESIZE macro automates the process of: - Identifying all character variables in a dataset - Calculating their actual maximum lengths - Generating and applying optimized LENGTH, FORMAT, and INFORMAT statements - Producing more efficient dataset without compromising data integrity By right-sizing variables, organizations can significantly reduce file sizes and processing times. In one example, a dataset with 4 million records and 118 variables saw a reduction in processing time from 51 minutes to just 4 minutes, along with notable savings in storage. This strategy improves performance, resource utilization, and overall usability of SAS datasets—making it a valuable data management approach in high-volume, real-world environments.

Divya Gadde

Managing large, raw datasets in SAS—especially those lacking metadata or proper layout documentation—often leads to inefficient data structures, inflated storage requirements, and slow performance. This paper presents a practical and scalable solution using the %RESIZE macro, designed to optimize character variable lengths by dynamically adjusting them to their actual maximum content length. The %RESIZE macro automates the process of: - Identifying all character variables in a dataset - Calculating their actual maximum lengths - Generating and applying optimized LENGTH, FORMAT, and INFORMAT statements - Producing more efficient dataset without compromising data integrity By right-sizing variables, organizations can significantly reduce file sizes and processing times. In one example, a dataset with 4 million records and 118 variables saw a reduction in processing time from 51 minutes to just 4 minutes, along with notable savings in storage. This strategy improves performance, resource utilization, and overall usability of SAS datasets—making it a valuable data management approach in high-volume, real-world environments.