Statistical procedures use ODS Graphics to create graphs as part of their output. ODS Graphics is described in detail in Chapter 24, Statistical Graphics Using ODS.
Before you create graphs, ODS Graphics must be enabled (for example, by specifying the ODS GRAPHICS ON statement). For more information about enabling and disabling ODS Graphics, see the section Enabling and Disabling ODS Graphics in Chapter 24, Statistical Graphics Using ODS.
The overall appearance of graphs is controlled by ODS styles. Styles and other aspects of using ODS Graphics are discussed in the section A Primer on ODS Statistical Graphics in Chapter 24, Statistical Graphics Using ODS.
When ODS Graphics is enabled, then for particular models the GLM procedure produces default graphics as follows:
If you specify a one-way analysis of variance model, with just one CLASS variable, the GLM procedure produces a grouped box plot of the response values versus the CLASS levels. For an example of the box plot, see the section One-Way Layout with Means Comparisons in Chapter 29, The ANOVA Procedure. Outliers are labeled by the observation number within the entire data set, including observations that are not used in the analysis. Therefore, the labels of the outliers in this box plot might differ from those in the box plot that is produced by the MEANS statement.
If you specify a two-way analysis of variance model, with just two CLASS variables, the GLM procedure produces an interaction plot of the response values, with horizontal position representing one CLASS variable and marker style representing the other, and with predicted response values connected by lines that represent the two-way analysis. For an example of the interaction plot, see the section PROC GLM for Unbalanced ANOVA.
If you specify a model that has two CLASS variables, and one variable is nested within the other, then the GLM procedure produces a nested box plot of the response values, where horizontal position represents one CLASS variable nested within the other CLASS variable.
If you specify a model that has a single continuous predictor, the GLM procedure produces a fit plot of the response values versus the covariate values, with a curve representing the fitted relationship. For an example of the fit plot, see the section PROC GLM for Quadratic Least Squares Regression.
If you specify a model that has a two continuous predictors and no CLASS variables, the GLM procedure produces a panel of fit plots as in the single predictor case, with a plot of the response values versus one of the covariates at each of several values of the other covariate.
If you specify an analysis of covariance model that has one or two CLASS variables and one continuous variable, the GLM procedure produces an analysis-of-covariance plot of the response values versus the covariate values, with lines representing the fitted relationship within each classification level. For an example of the analysis of covariance plot, see Example 53.4.
If you specify an LSMEANS statement with the PDIFF option, the GLM procedure produces a plot appropriate for the type of LS-means comparison. For PDIFF=ALL (which is the default if you specify only PDIFF), the procedure produces a diffogram, which displays all pairwise LS-means differences and their significance. The display is also known as a "mean-mean scatter plot" (Hsu 1996). For PDIFF=CONTROL, the procedure produces a display of each noncontrol LS-mean compared to the control LS-mean, with two-sided confidence intervals for the comparison. For PDIFF=CONTROLL and PDIFF=CONTROLU, a similar display is produced, but with one-sided confidence intervals. Finally, for the PDIFF=ANOM option, the procedure produces an analysis-of-means plot, which compares each LS-mean to the average LS-mean.
If you specify a MEANS statement, the GLM procedure produces a grouped box plot of the response values versus the effect for which means are being calculated. Outliers are labeled by the observation number within only the data that have no missing values, thus excluding observations that are not used in the analysis. Therefore, the labels of the outliers in this plot might differ from the labels of outliers in the box plot that is the default for one-way models.
In addition to the default graphics mentioned previously, you can request plots that help you diagnose the quality of the fitted model.
The PLOTS=DIAGNOSTICS option in the PROC GLM statement requests that a panel of summary diagnostics for the fit be displayed. The panel displays scatter plots of residuals, absolute residuals, studentized residuals, and observed responses by predicted values; studentized residuals by leverage; Cook’s D by observation; a Q-Q plot of residuals; a residual histogram; and a residual-fit spread plot.
The PLOTS=RESIDUALS option in the PROC GLM statement requests scatter plots of the residuals against each continuous covariate.
PROC GLM assigns a name to each graph that it creates using ODS. You can use these names to refer to the graphs when you use ODS. The names are listed in Table 15.
ODS Graphics must be enabled before requesting plots. For more information about ODS Graphics, see Chapter 24, Statistical Graphics Using ODS.
Table 15: Graphs Produced by PROC GLM
| ODS Graph Name | Plot Description | Option |
|---|---|---|
| ANCOVAPlot | Analysis-of-covariance plot | Analysis-of-covariance model |
| ANOMPlot | Plot of LS-mean differences against average LS-mean | LSMEANS / PDIFF=ANOM |
| BoxPlot | Box plot of group means | One-way ANOVA model or MEANS statement |
| ContourFit | Plot of predicted response surface | Two-predictor response surface model |
| ControlPlot | Plot of LS-mean differences against a control level | LSMEANS / PDIFF=CONTROL |
| DiagnosticsPanel | Panel of summary diagnostics for the fit | PLOTS=DIAGNOSTICS |
| CooksDPlot | Cook’s D plot | PLOTS=DIAGNOSTICS(UNPACK) |
| ObservedByPredicted | Observed by predicted | PLOTS=DIAGNOSTICS(UNPACK) |
| QQPlot | Residual Q-Q plot | PLOTS=DIAGNOSTICS(UNPACK) |
| ResidualByPredicted | Residual by predicted values | PLOTS=DIAGNOSTICS(UNPACK) |
| ResidualHistogram | Residual histogram | PLOTS=DIAGNOSTICS(UNPACK) |
| RFPlot | RF plot | PLOTS=DIAGNOSTICS(UNPACK) |
| RStudentByPredicted | Studentized residuals by predicted | PLOTS=DIAGNOSTICS(UNPACK) |
| RStudentByLeverage | RStudent by hat diagonals | PLOTS=DIAGNOSTICS(UNPACK) |
| DiffPlot | Plot of LS-mean pairwise differences | LSMEANS / PDIFF |
| FitPlot | Plot of predicted response by predictor | Model with one continuous predictor |
| IntPlot | Interaction plot | Two-way ANOVA model |
| LinesPlot | Plot of mean or LS-mean comparison lines | Lines comparison options in LSMEANS or MEANS |
| NestPlot | Nested box plot | Model in which one classification predictor is nested within the other |
| ResidualPlots | Plots of the residuals against each continuous covariate | PLOTS=RESIDUALS |