Graphics Reference
In-Depth Information
Table
.
.
he superindicator matrix representation (Gifi,
) of a categorical data set
Sleeping Bag
Price
Fiber
Quality
n e K i l o B a g
S u n d
Kompakt Basic
F i n a r k T o u r
I n t e r l i g h t L y x
K o p a k t
Touch the Cloud
C a t 's e o w
I g l o o S u p e r
Donna
Tyin
T r a v e l l e r s D r e a m
Y e t i L i g h t
C l i b e r
V i k i n g
E i g e r
Climber light
C o b r a
C o b r a C o f o r t
F o x fi r e
o n t B l a n c
of edges, identify the positions of the nodes in some space and calculate the curves
that connect them. Hence, in order to draw a graph, one has to make the following
two choices: (i) selection of the space and (ii) selection of the curves. For example,
grid layouts position the nodes at points with integer coordinates, while hyperbolic
layoutsembedthepointsonasphere.Mostgraphdrawingtechniquesusestraight
lines between connected nodes, but some use curves of a certain degree (Di Battista
et al.,
).
Manylayoutalgorithms arebased onasetof aesthetic rulesthatthedrawingneeds
toadhereto.Popularrulesarethatnodesandedgesmustbeevenly distributed,edges
should have similar lengths, edge crossings must be kept to a minimum, etc. Some
of these rules are important in certain application areas. Further, many of these rules
lead toa corresponding optimization problem,albeit intractable in certain cases. For
example, the edge-crossing minimization is provably NP-hard and hence computa-
tionally intractable (Di Battista et al.,
). In many cases, a basic layout is obtained
by a computationally fast algorithm, and the resulting drawing is postprocessed to
