SAS is very popular in North America but is not dominate in quantitative analytic field. Matlab, EXCEL VBA and C++ are employed widely. SAS is preferred when dealing with large datasets, or building models with well-established procedures. However, people prefer C++ for package with flexible industrial application, prefer Matlab with complicated matrix manipulation, especially under high dimensional calculations, prefer EXCEL for graphic application. Therefore, SAS is a must but not unique tool in the future and it has own shortcomings. For new statistical application, R, similarly to S-plus has its own advantage as well. Actually, information exchanging among softwares are developing and has some progress.
Theoretical study on the underlying methodology is more important than the software application themselves. Some calculation in SAS is not optimized or accurate. External calculation out of some procedure is required. This is the solid base to push SAS to be a better software. Misunderstanding SAS output is terrible.


雷达卡
京公网安备 11010802022788号







