SAS Visual Statistics 8.5 is now available!

[mvgilliland]

SAS Visual Statistics 8.5 (on SAS Viya 3.5) is now available, with many new capabilities including:

Graphical User Interface

• A confusion matrix that displays the classification results for categorical response models is available.
• Several new assessment statistics are available for all models of the same type.
• You can now derive predicted items from generalized additive models and nonparametric logistic regression models.
• You can now export and save generalized additive models and nonparametric logistic regression models.
• In the cluster object, a details table has been added for the parallel coordinates plot.
• The Bin method property that specifies the way that measure predictors are binned has been added to the decision tree object.
• There is a new option for autotuning decision tree models: Leaf size.
• In the settings, you can now specify to automatically convert measure variables with two levels to category variables when a data set is first opened in SAS Visual Analytics.
• In the settings, you can now set a default assessment statistic to use in both continuous and classification models.

Procedures and Actions

• You can now perform generalized additive model selection by using PROC GAMSELECT or the gamSelect action in the gam action set.
• You can now perform nonnegative matrix factorization by using PROC NMF or the nmf action in the nmf action set.
• To perform simulation from systems of distributions defined by input moments, you can now use PROC SIMSYSTEM or the simsk action in the simSystem action set.
• You can now perform sparse linear regression with sandwich variance estimation by using PROC SANDWICH or the sparseReg action in the sandwich action set.
• To fit a nonlinear model that you provide as a CMP program, you can now use the nlmod action in the nonlinear action set.
• You can now perform elastic net selection by using PROC REGSELECT or the glm action in the regression action set.
• A factor analytic covariance type is now available in PROC MBC and the mbcfit action in the mbc action set.
• You can now perform postfit scoring by using PROC LMIXED with PROC ASTORE.

Find full details on the SAS Visual Statistics product page.