Display Attributes

Overview

The display attributes for the lines, colors, marker symbols, and text used in a graph are derived from the ODS style that is in effect when the graph is produced. These display attributes might also be influenced by grouped data. To override default display attributes, all GTL plot statements provide options that manage the graph’s visual appearance. For example, a BOXPLOT statement provides an OUTLIERATTRS= option that manages the visual appearance of outliers.

Two ways are generally available for modifying a graph’s display attributes:

Change the ODS style that is in effect for the graph. ODS Styles provides an overview of the use of styles in a graph. SAS Graph Template Language: User's Guide discusses the use of styles in more detail.
Override default style settings using GTL statement options. Some examples are given in the sections that follow.

Display Attributes for Non-Grouped Data

Display Attributes documents the attribute settings that can be specified for the lines, data markers, text, or area fills in a plot. The defaults for these attributes are defined on style elements, but you can use attribute options on the plot statement to change the defaults.

For example, the LINEPARM statement provides a LINEATTRS= option that specifies the color, line pattern, or line thickness of the plot line. For non-grouped data, if you do not set a line pattern in your template, then the default line pattern for the plot is obtained from the GraphDataDefault:LineStyle style reference.

To change the default line pattern, a PATTERN= suboption on LINEATTRS= is available. "Common Line Patterns" shows the most common line patterns available for the PATTERN= suboption.

"Common Line Patterns"

the left column shows the names for the line patterns
the center column illustrates the type of line the name requests
the right column shows the SAS line-style numbers for the line patterns

Available Line Patterns provides the complete list of line patterns that can be used with the GTL.

In the following template definition, the LINEPARM statement’s LINEATTRS= option overrides the default line pattern by specifying PATTERN=DASH:

proc template;
  define statgraph patternchange;
    begingraph;
      layout overlay;
        scatterplot y=height x=weight;
        lineparm yintercept=intercept  slope=slope /
          lineattrs=(pattern=dash);
      endlayout;
    endgraph;
  end;

Other display options can be managed the same way. For example, the SCATTERPLOT statement provides a MARKERATTRS= option that specifies the color, size, symbol, and weight of the plot data markers. For non-grouped data, if you do not set a marker symbol in your template, then the default marker symbol is obtained from the GraphDataDefault:MarkerSymbol style reference.

To change the default marker symbol, a SYMBOL= suboption on MARKERATTRS= is available. "Marker Symbols" shows the marker symbols available for the SYMBOL= suboption.

"Marker Symbols"

In the following template definition, the SCATTERPLOT statement’s MARKERATTRS= option overrides the default marker symbol by specifying SYMBOL=CIRCLEFILLED, which uses a filled circle to represent the data points.

proc template;
  define statgraph symbolchange;
    begingraph;
      layout overlay;
        scatterplot y=height x=weight /
          markerattrs=(symbol=circlefilled);
      endlayout;
    endgraph;
  end;

Display Attributes for Grouped Data

Display Attributes documents the attribute settings that you can specify for the lines, data markers, text, or area fills in a plot. For grouped data (that is, when you use the GROUP= option in a plot statement), each distinct group value can be represented in the graph by a different combination of line pattern, fill pattern, color, and marker symbol (depending on the graph type). The defaults for these features are set by the LineStyle, Color, ContrastColor, FillPattern, and MarkerSymbol attributes of the GraphData1–GraphDataN style elements.

Note: The MarkerSize and LineThickness style attributes are not honored in the case of grouped data.

"Common Line Patterns" shows the common line patterns available, and "Marker Symbols" shows the marker symbols available.

For grouped plots, the style in effect and the plot settings determine which line patterns, area fills, and plot symbols are used. If different line patterns, colors, or marker symbols are used to represent group values, then the style determines the sequences of the line patterns, colors, or marker symbols that are used for the group values. (As discussed in Cycling through Group Attributes in Overlaid Plots, other plot settings might also influence the sequence.) If the number of group values exceeds the number of style elements, the following occurs for the subsequent group values:

The sequence of line patterns and plot markers that are specified by the style is repeated.
New fill colors are generated by repeating the GraphData1–GraphDataN colors and varying the shade of each original color on each cycle. The shade variations alternate between one shade lighter and one shade darker on each cycle.

You can use attribute options on the plot statement to change the default display attributes used for group data. For example, in the following template definition, the LINEPARM statement’s LINEATTRS= option specifies PATTERN=DASH. This explicit setting overrides the default line pattern for the plot lines and uses dashed lines for all of the plots, leaving color to distinguish among group values.

proc template;
  define statgraph dashedline;
    begingraph;
      layout overlay;
        scatterplot y=height x=weight / group=gender;
        lineparm yintercept=intercept  slope=slope / group=gender
          lineattrs=(pattern=dash);
      endlayout;
    endgraph;
  end;

Rather than setting the same line pattern on all group values, you can change the default sequence of line patterns that is used for grouped values. To do so, set the LineStyle attribute in some of the style elements GraphData1 through GraphDataN.

In the following example, a style is defined to change the default line pattern for the first two lines in the pattern sequence. In this example, the style is derived from the DEFAULT style, which is available for the HTML destination. Values are set for the LineStyle attributes in the GraphData1 and GraphData2 style elements. The first default line in the sequence has long dashes (style value 6) and the second line has short dashes (style value 4). The LineStyle settings for the remaining GraphData elements are not set, so are derived from the parent style (DEFAULT). This new line sequence is used as the default line sequence for any plot that uses the MyDefault style. To apply the style to a graph, the STYLE= option is used in the ODS HTML statement to specify the style name.

Here is the code for this example.

/* Sort the SASHELP.CLASS data by sex and age. */
proc sort data=sashelp.class(keep=height weight sex age)
     out=class;
  by sex age;
run;

/* Generate slope and intercept data for plot reference lines. */
proc robustreg data=class method=m
      outest=stats(rename=(weight=slope));
   by sex;
   model height=weight;
run;

data class;
   merge class stats(keep=intercept slope sex);
run;

proc template;
   /* Create custom style STYLES.MYDEFAULT from the STYLES.DEFAULT style. */
   define style Styles.MyDefault;
      parent=Styles.Default;
      style GraphData1 from GraphData1 /
         LineStyle=6;
      style GraphData2 from GraphData2 /
         LineStyle=4;
   end;

   /* Create the plot template. */
   define statgraph testPattern;
      begingraph;
         layout overlay;
            scatterplot y=height x=weight / group=sex;
            lineparm x=0 y=intercept slope=slope / group=sex name="lines";
            discretelegend "lines";
         endlayout;
      endgraph;
    end;
run;

/* Generate the plot. */
ods _all_ close;
ods html style=MyDefault; /* Apply style MyDefault to the graph. */

proc sgrender data=class template=testPattern;
run;

Similarly, for grouped data, you can set the MarkerSymbol attribute in each of the style elements GraphData1 through GraphDataN. In the following example, a style is defined to change the default sequence that is used for the first three marker symbols in grouped plots. Values are set for the MarkerSymbol attributes in the GraphData1 through GraphData3 style elements. This new sequence is used as the default marker symbol sequence for any plot that uses the MyDefault style.

Here is the code for this example.

Note: The data that was generated in the previous example is used again in this example.

proc template;
   /* Create custom style STYLES.MYDEFAULT from the STYLES.DEFAULT style. */
   define style Styles.MyDefault;
      parent=Styles.Default;
      style GraphData1 from GraphData1 /
         MarkerSymbol="DIAMOND";
      style GraphData2 from GraphData2 /
         MarkerSymbol="CROSS";
      style GraphData3 from GraphData3 /
         MarkerSymbol="CIRCLE";
   end;

   /* Create the plot template. */
   define statgraph testSymbols;
      begingraph;
         layout Overlay;
            scatterPlot y=height x=weight / group=age name="symbols";
            discretelegend "symbols" / title="Age";
         endlayout;
      endgraph;
   end;
run;

/* Generate the plot. */
ods html close;
ods html style=MyDefault; /* Apply style MyDefault to the graph. */

proc sgrender data=class template=testSymbols;
run;

Cycling through Group Attributes in Overlaid Plots

Overlay-type layouts provide the CYCLEATTRS= options that specifies whether the default visual attributes of lines, marker symbols, and area fills in nested plot statements automatically change from plot to plot. When CYCLEATTRS=TRUE, all applicable plot statements (SCATTERPLOT, SERIESPLOT, and others) are sequentially assigned the next unused GraphDataN style element. (The sequence is overridden for plot statements that have an explicit setting, either through a style element assignment or option settings.) No plot retains its default (implicit) style element.

In the following example, assuming ungrouped data, the series plots are assigned line properties based on the GraphData1, GraphData2, and GraphData3 style elements. The reference line uses GraphReference, not GraphData4.

layout overlay / cycleattrs=true;
  seriesplot x=date y=var1;
  seriesplot x=date y=var2; 
  seriesplot x=date y=var3;
  referenceline x=cutoff / lineattrs=GraphReference;
endlayout;

If one of the plots in this example uses grouped data, the grouped plots also participate in the default cycles. For example, if the second plot has three groups, it generates three plots, which are assigned line properties based on the GraphData2, GraphData3, and GraphData4 style elements.

If the plot statement that uses grouped data also uses the INDEX= option to manage the group values (see Remapping Groups for Grouped Data), the INDEX= option overrides the default behavior. In that case, the grouped plots do not participate in the default cycling.

When one or more of the plots within the layout override the default cycling behavior, the arrangement of the plots within the layout might affect the default mapping of the GraphDataN elements to those statements that participate in the default cycling.

Remapping Groups for Grouped Data

Indexing can be used to collapse the number of groups that are represented in a graph. For example, if 10 groups are in the data, indexes 1 and 2 can be assigned to the first two groups, and index 3 can be assigned to all other groups. The third through tenth data groups are treated as a single group in the graph.

Indexing can control the order in which colors, area fills, marker symbols, and line styles are mapped to group values in a graph. This ordering method is needed only for coordinating the data display of multiple graphs when the default mapping would cause group values to be mismatched between graphs.

For example, consider two studies of three drugs, A, B, and C. If Study 1 uses all three drugs, then the first combination of color and marker symbol is mapped to Drug A. The second combination of color and marker symbol is mapped to Drug B, and the third is mapped to Drug C. If Study 2 omits Drug A, then the first combination of color and marker symbol is mapped to Drug B, and the second is mapped to Drug C. If the two graphs are viewed together, then this default mapping causes the group values to be mismatched. The visual attributes that represent Drug A in the first graph represent Drug B in the second graph. Those that represent Drug B in the first graph represent Drug C in the second group.

The GROUP= option mappings can be made consistent between the two graphs by creating an index column for each study. For these example studies, the GROUP and INDEX columns are the following:

Study 1

Drug1	Index1
A	1
B	2
B	2
C	3

Study 2

Drug2	Index2
B	2
C	3
C	3

If the graph for Study 1 specifies INDEX=INDEX1 and the graph for Study 2 specifies INDEX=INDEX2, then the second combination of color and marker symbol is mapped to Drug B in both graphs. The third combination of color and marker symbol is mapped to Drug C in both graphs.

Interactions between Options

When you use GTL statement options to manage the graph display, interactions between options might cause some option settings to be ignored. For example, an ENTRYTITLE statement provides BORDER= and BORDERATTRS= options for managing a border line around the graph title. Border attributes that are set on the BORDERATTRS= option have no effect on the graph title unless the title border line is displayed by setting BORDER=TRUE.

Similarly, if a BOXPLOT statement’s DISPLAY= option suppresses the display of outliers in a box plot, then using the OUTLIERATTRS= option to set outlier attributes has no effect. The OUTLIERATTRS= settings only take effect if DISPLAY= enables the display of outliers.

The option interactions are not limited to options that simply manage visual elements. For example, on a BOXPLOT, if the EXTREME= option extends the box whiskers beyond the fences, then outliers are suppressed in the plot and options that affect the outliers are ignored, if set.

The documentation for each GTL statement identifies the option interactions that might occur on that statement.