Graphics Reference
In-Depth Information
linearizing the embedded vectors or tensors. We also assume that a data set consists
of one or more of such data items/records and that for each position in the vector we
can calculate a minimum and maximum value. his allows us to normalize the data
to facilitate mapping to graphical attributes.
Variables/dimensions can be independent or dependent, which might imply that
some ordering or grouping of dimensions could be beneficial. hey can be of homo-
geneous type,such as a set of exam grades, or of mixed/heterogeneous types, such as
mostcensus data. hismightsuggesttheuseofaconsistent mapping(e.g.,alldimen-
sions map to line lengths) or separation based on type so that each distinct group of
related dimensions might control one type of mapping.
Mappings
7.3
Manyauthors have developed lists of graphical attributes towhichdata values can be
mapped (Cleveland and McGill,
; Cleveland,
; Bertin,
). hese include
position (
-,
-, or
-D), size (length, area, or volume), shape, orientation, material
(hue, saturation, intensity, texture, or opacity), line style (width, dashes, or tapers),
and dynamics (speed of motion, direction of motion, rate of flashing).
Using these attributes, a wide range of possible mappings for data glyphs are pos-
sible. Mappings can be classified as follows:
One-to-one mappings, where each data attribute maps to a distinct and different
graphical attribute;
One-to-many mappings, where redundant mappings are used to improve the ac-
curacy and ease at which a user can interpret data values; and
Many-to-one mappings, where several or all data attributes map to a common
type of graphical attribute, separated in space, orientation, or other transforma-
tion.
One-to-one mappings are oten designed in such a way as to take advantage of the
user's domain knowledge, using intuitive pairings of data to graphical attribute to
ease the learning process. Examples include mapping color to temperature and flow
direction to line orientation. Redundant mappings can be useful in situations where
the number of data dimensions is low and the desire is to reduce the possibility of
misinterpretation. For example, one might map population to both size and color to
ease analysis for color-impaired users and facilitate comparison of two populations
with similar values. Many-to-one mappings are best used in situations where it is
important to not only compare values of the same dimension for separate records,
but also compare different dimensions for the same record. For example, mapping
each dimension to the height of a vertical bar facilitates both intrarecord and inter-
record comparison. In this paper, we focus primarily on one-to-one and many-to-
one mappings, although many of the principles discussed can be applied to other
mappings.
