SAS/STAT vs SAS Visual Statistics

SAS/STAT is statistical analysis software for specialized and enterprise-wide analytics. Its library includes more than 90 procedures for methods such as regression, survey and survival analysis, multiple imputation and Bayesian modeling. Other documented areas include ANOVA, categorical, cluster, multivariate and nonparametric analysis, as well as propensity scores and causal treatment-effect estimation. Methods are available for small data sets, missing values and large data tasks. SAS describes 14 SAS/STAT procedures as multithreaded and makes 12 SAS High-Performance Statistics procedures available for single-machine use. Predictive modeling options include regression, generalized additive, mixed and nonlinear models, and decision trees. Hundreds of statistical graphs and customizable graphics are included. SAS says the software runs on major computing platforms, accesses nearly any data source and integrates into computing environments. A free trial provides access to SAS Viya in a private trial environment. Pricing is available on request. SAS lists SAS/STAT on the SAS Viya platform and SAS/STAT 15.4 among its most recent releases.

SAS Visual Statistics is a statistical visualization and modeling product on SAS Viya. It supports analysis from raw data toward strategic action, with visual model building, variable adjustment, and outlier removal; users can see how those changes affect outcomes. Multiple users can collaborate on the same data model, and the environment supports Python, R, or SAS. Listed techniques include clustering, decision trees, linear and logistic regression, generalized linear and additive models, and nonparametric logistic regression.

Imports include SAS data sets, Excel, delimited text, CSV, DTA, JMP, Parquet, and SPSS. Programmers can access the Viya CAS server through SAS Studio, PROC CAS, Python, R, Lua, Java, and public REST APIs. Native Python Pandas DataFrame integration supports sending DataFrames to CAS and retrieving results for use with packages such as Pandas, matplotlib, Plotly, and Bokeh. SAS Cloud Analytic Services processes data in memory and distributes work across cluster nodes. The product is cloud-deployed, paid, and priced on request; no free plan is listed, but a free trial is available.