Key Principles of Data Analytics using SAS

Based on our decades of SAS experience, we want to share with you some of the key principles. As there are exceptions to any rule, you may find for these as well.

  • Incorrect result can erode credibility and trust built over many analyses and years. Hence validation of data analysis results is critical. One of the first steps of validation is checking SAS log. Even do not ignore warning written in log.
  • SAS works on principal of sequential processing especially SAS data steps . Each observation from the input dataset will be processed/executed for the steps mentioned in SAS codes one by one. Unless instructed specifically, SAS does not retained values from previous observation(s).
  • Each of the SAS statement is ended by semi-comma and a single statement can be across multiple lines. This functionality can be really useful to make the SAS code easy to read for users. But missing a semi-comma is one of the most common mistake by SAS users.
  • Also, one of the most favorable SAS functionalities for SAS programmers or Data Analysts is that SAS code is not case sensitive. Data Analyst can make SAS code look better or easy of read for users by combining Lower and Upper case, but that does not any difference from SAS execution point of view.
  • Due to a huge list of functions, Procs and SAS data steps flexibility, one can do almost anything required for a Data Analysis and also number of different ways. Still the results with be same.
  • Last but not the least point from my side is that by default an observation with missing value will be excluded from processing. For example, if Gender takes missing value, in printing count and percentage, these observations will be ignored.

Leave a Comment