Graphics Reference
In-Depth Information
FIGUREĀ 5-29 Highlighting in the background
However, wherever your highlighting fits in the hierarchy, be sure the new visual
cues don't conflict with existing ones. If you have a bar chart that uses length
as a visual cue, you obviously do not highlight with length, too. Have a scatter
plot? Don't highlight with position. Heat map? Highlight with the color palette
rather than introduce hues that change visual patterns.
Ask yourself how people decode information via shapes and
colors in a visualization, and then don't get in the way. For
example, in Figure 5-30 on the left is a bar graph with no
highlighting, and the charts on the right show unsuccessful
attempts to highlight. Why don't they work? A bar chart uses
length as its visual cue, so when you extend a bar, the new
length changes the value. Change width, and the new bar
fills more area. (Bar charts actually use area to encode data,
but because width stays constant, you can decode values via
bar height.) Then on the far right: A shift up doesn't exactly
change the value, but it makes the chart less readable.
Note: These conflicting visual cues are in the
context of how you use bar charts, but you of
course must consider conflicts within the context
of your own visualization and how you encode
your data.
In contrast, Figure 5-31 shows highlighting with visual cues
not used by the bar graph. The color, border, and pointer
send focus to the bar of interest but don't change the overall
visual pattern.
Note: Highlight with unused visual cues.
Otherwise, you change perceived patterns
and make it more difficult to interpret the
visualization.
Search WWH ::




Custom Search