SAS Viya uses SAS Cloud Analytic Services (CAS) to perform tasks. The smallest unit of work for the CAS server is a CAS action. CAS actions can load data, transform data, compute statistics, perform analytics, and create output. Each action is configured by specifying a set of input parameters. Running a CAS action in the CAS server processes the action's parameters and the data and creates an action result. For more information, see "Servers and Sessions" in An Introduction to SAS Viya Programming.
SAS Viya provides a variety of interfaces for running CAS actions, including the following:
- from a SAS session using the CAS procedure. The CAS procedure uses the CASL language for specifying CAS actions and their input parameters. CASL also supports normal program logic such as conditional and looping statements and user-written functions.
- from Python or Lua using the SAS Scripting Wrapper for Analytics Transfer (SWAT) libraries.
- from Java using the CASClient class.
- via REST using the CAS REST APIs.
CAS actions are organized into action sets. Each action set defines an application programming interface. SAS Viya currently provides the following action sets:
- Statistics action sets
provide summary statistics, clustering, regression, sampling, principal, component analysis, and more.
- Analytics action sets
provide text and numeric analytics.
- System action sets
run SAS code via DATA step or DS2, manage CAS libraries (caslibs) and tables, manage CAS servers and sessions, and more.
- Data Mining and Machine Learning action sets
support scoring, factorization, neural networks, support vector machines, graph/network analysis, text mining, and more.