The SGPANEL Procedure

Concepts

The SGPANEL procedure has a required PANELBY statement that is used to define the classifier variables for the panel. This statement must be specified before any plot, axis, or legend statement or else an error occurs. You can use options in the PANELBY statement to control the attributes of the panel. For example, you can use the COLUMNS= option to specify the number of columns in the panel.

SGPANEL can use four different layouts, which are specified by the LAYOUT= option in the PANELBY statement. The layout determines how your classifier variables are used to create the panel, and also affects the number of classifier variables that you can specify.

The default layout is PANEL. With this layout, you can specify any number of classifier variables. The graph cells in the panel are arranged automatically, and the classifier values are displayed above each graph cell in the panel. When you specify multiple classifier variables, the order of the classifier variables determines how the graph cells are sorted.

Example of the PANEL Layout shows an example of the PANEL layout.

Example of the PANEL Layout

[Panel Layout Style]

Another layout is called LATTICE. This layout requires exactly two classifier variables. The values of the first variable are assigned as columns, and the values of the second variable are assigned as rows. The classifier values are displayed above the columns and to the right side of the rows.

Example of the LATTICE Layout shows an example of the LATTICE layout.

Example of the LATTICE Layout

[Lattice Layout Style]

Two additional layouts are available, which are called COLUMNLATTICE and ROWLATTICE. These layouts require exactly one classifier variable. The values of the classifier variable are assigned as cells in a single row or column.

Note: The COLUMNLATTICE and ROWLATTICE layouts are available with SAS 9.2 Phase 2 and later. [cautionend]

Example of the COLUMNLATTICE Layout shows an example of the COLUMNLATTICE layout.

Example of the COLUMNLATTICE Layout

[COLUMNLATTICE Layout Style]

If you have a large number of classifier variables, then the best method for creating a panel is to choose one or two classifiers for the PANELBY statement and specify the remaining variables in a BY statement. This method maximizes the space for the plots and generates results that are easier to interpret.

Plot Content

Each graph cell in your panel contains one or more plots, and there are four basic types of plots that you can create with the SGPANEL procedure:

Basic plots: scatter, series, step, band, needle plots, and vector plots
Fit and confidence plots: loess, regression, and penalized B-spline curves
Distribution plots: box plots, histograms, normal density curves, and kernel density estimates
Categorization plots: dot plots, bar charts, and line plots

Not all of the plot types can be used together in the same PROC SGPANEL step. The following table shows which of the plot types can be used together:

Plot Type Compatibility
	Basic	Fit and Confidence	Distribution	Categorization
Basic	x	x
Fit and Confidence	x	x
Distribution			x
Categorization				x

Note: Box plots cannot be combined with any other plot types. [cautionend]

If you submit a PROC SGPANEL step that combines two incompatible plot statements, then an error appears in the log.

The SGPANEL procedure draws the plots in your graph in the same order that you specify the plot statements. Because of this, it is important to consider the order of your plot statements so that your plots do not obscure one another. For example, if you specify a BAND statement after a SCATTER statement, then the band plot might obscure the markers in your scatter plot. You can also avoid obscuring your data by using the TRANSPARENCY= option to make your plots partially transparent.

Plot Axes

The SGPANEL procedure contains two statements that enable you to change the type and appearance for the axes of the graph cells in your panel: COLAXIS and ROWAXIS.

By default, the type of each axis is determined by the types of plots that use the axis and the data that is applied to the axis.

The SGPANEL procedure supports the following axis types:

Discrete: The axis contains independent data values rather than a range of numeric values. Each distinct value is represented by a tick mark. Discrete is the default axis type for character data.
Linear: The axis contains a linear range of numeric values. Linear is the default axis type for numeric data.
Logarithmic: The axis contains a logarithmic range of values. The logarithmic axis type is not used as a default.
Time: The axis contains a range of time values. Time is the default axis type for data that uses a SAS time, date, or datetime format.

Some types of plot do not support all of the axis types. For example, needle plots cannot use a discrete vertical axis. See the documentation for each plot statement to determine whether any axis type restrictions apply.

Panel Legends

The SGPANEL procedure creates a legend automatically based on the plot statements and options that you specify. The automatic legend functionality determines which information is likely to be useful in the legend. You can override this behavior by defining your own legend with the KEYLEGEND statement or by specifying the NOAUTOLEGEND option in the PROC SGPANEL statement.

You can specify the labels that represent your plots in the legend by using the LEGENDLABEL= option in the corresponding plot statements.

You can create customized legends by using one or more KEYLEGEND statements. You can specify which plot statements are assigned to the legend, and use options to control the title, location, and border of the legend. For more information, see KEYLEGEND Statement.

Automatic Differentiation of Visual Attributes

Depending on the plots and options that you specify, the SGPANEL procedure can automatically assign different style attributes to the plots in your graph. For example, if you specify two series plots, then each series plot automatically uses a different line pattern and line color by default. If different attributes are not assigned by default, you can force the procedure to assign different style attributes by using the CYCLEATTRS option in the PROC SGPANEL statement. You can also disable automatic attribute differentiation by using the NOCYCLEATTRS option in the PROC SGPANEL statement.

Units of Measurement

Some options such as LINEATTRS enable you specify the unit of measurement as part of the value. The following table contains the units that are available:

Measurement Units
Unit	Description
CM	centimeters
IN	inches
MM	millimeters
PCT or %	percentage
PT	point size, calculated at 100 dots per inch
PX	pixels

Marker Symbols

The MARKERATTRS= option in some of the plot statements enables you to specify the marker symbol that is used to represent your data. List of Marker Symbols shows the marker symbols that you can use.

List of Marker Symbols

[Marker Symbols]

Line Patterns

The LINEATTRS= option in some plot statements enables you to specify the line pattern that is used for the lines in your plot. List of Line Patterns shows the line patterns that you can use.

List of Line Patterns

[List of Line Patterns]

Top of Page