Gallery

This is an illustrated gallery of plots and graphs found in SAS Visual Statistics.
Tip
Use the following images for orientation. Actual appearance and functionality are affected by the underlying data, any styles that you apply, and the interface that you are using.
Fit Summary window
Displays the p-value of each modeling variable on a log scale. The alpha value, plotted as -log(alpha), is shown as a vertical line that you can click and drag to adjust. A histogram of the p-values is displayed at the bottom of the window.
This window is divided when a group by variable is used. The left side lists the groups and the right side condenses the p-values for each group into a single linear scatter plot. You can click on a group on the left side to change the Residual, Influence, and Assessment plots to show only the results for that group.
Residual Plot
Displays various residual plots for the model. When the plot labels are buttons, you can select the values that are plotted on that axis. Each model has a unique set of plot combinations available.
Assessment window
For measure target variables, Assessment plots the average predicted and observed values against the binned data set. For category target variables, it provides the Lift, ROC, and Misclassification plots.
Influence Plot
Plots each observation against various computed statistics. The X axis label is a button that enables you to determine what is plotted. Each model has a unique set of plot combinations available.
Decision Tree and Treemap
Displays the decision tree and the decision treemap. You can interactively train the decision tree from this window. Use the scroll wheel on your mouse to zoom in or out on the location of the mouse pointer.
Leaf StatisticsLeaf Statistics
Provides a stacked histogram of the response variable for each leaf node in the decision tree.
Cluster Matrix
Displays the two-dimensional projection of every cluster for each pair of modeling variables. To view a larger plot of an individual projection, right-click in that cell, and select Open.
Parallel Coordinates Plot
Displays a color-coded strand for each observation, initially sorted by cluster membership. You can restrict the display to observations that match specific clusters or ranges of the modeling variables.
Last updated: August 16, 2017