STATISTICA software has been recognized as a performance leader in data analysis applications (Click here for the record of published reviews in the past 10 years). However, none of the published reviews has evaluated the performance of STATISTICA using large data sets, and to the best of our knowledge, no such published benchmarks exist for the competing products, either.

Due to the dynamic growth of data mining applications in a wide variety of industries, customers are asking for information about the performance of STATISTICA software on large data sets.

The following benchmark tests were conducted using a representative selection of moderate to large data sets, and they compare the performance of STATISTICA Data Miner to software from SASŪ (a publisher of SAS Enterprise Miner, a data mining application that is (a) promoted as the most suitable program for handling large data sets, and (b) a program that appears closer to STATISTICA Data Miner in terms of its intended applications and audience, than any other product).

Testing methodology. All tests were performed on DELL workstation computers, 1.7 GHz (Pentium 4), with 256 Megabytes of RAM, and an IDE (7,200 RPM) disk drive. The data set for each test resided on the local hard drive. All tests were performed on a clean system, under Windows 2000 Professional. In all cases, the data were stored in the optimal (native) file format for SAS and STATISTICA, respectively, and they consisted of random, real numbers. Also, to avoid confounding these tests by differences in options, settings, etc., no filters, weights, or case selection conditions were used, and only simple descriptive statistics were computed (i.e., means, standard deviations, minimum and maximum values, and valid N). The computer was rebooted before the application software was loaded (and that application was the first one open after the rebooting). Each test on a given data file was performed five times (consecutively), to arrive at an average speed for each test. All reported times are in seconds. The computer was rebooted before testing each consecutive data file.

SAS is a registered trademark of SAS Institute.

Back to the STATISTICA Data Miner page.

2300 East 14th Street, Tulsa, OK 74104
Phone: (918) 749-1119; Fax: (918) 749-2217
e-mail: info@statsoft.com

©Copyright StatSoft, Inc., 1984-2004.
StatSoft, StatSoft logo, STATISTICA, SEWSS, SEDAS, Data Miner, SEPATH and GTrees are trademarks of StatSoft, Inc.