Graphics Reference
In-Depth Information
middle panel shows the median value for the variable used in sorting. Carr et al.
(
) exploit this in a special design for the
US states plus Washington, DC. In
other situations, it might be preferable to avoid creating groups with few regions.
he right column (Partitioning
) in Table
.
is also symmetric, but it avoids small
counts wherepossible.hevisualappealofsymmetry islostwhentherearetoomany
groups to display in one page or one screen image. In this case, it su
ces to fill all
panels except a single panel that contains the letover regions modulo the grouping
size. It does not seem to matter whether the partially filled panel appears at the top,
near the middle, or at the bottom of the page.
he micromaps in LM plots are intended to be caricatures that generally serve
three purposes. he first purpose is to provide a recognizable map. Boundary details
are not important here. Boundaries only need to communicate region identity and
neighbor relationships.
he second purpose is to provide a visible color link. Tiny areas do not support
easy color identification. As indicated above, early micromaps for states adapted the
hand-drawn state visibility map of Monmonier (
). his map simplified bound-
aries and increased the size of small US states suchas RhodeIsland. he EPA Cumu-
lative Exposure Project (CEP) Web site (Symanzik et al.,
) made it necessary to
develop an algorithm to generalize boundaries for US states and US counties. Most
of the Web-based micromaps described in Sect.
.
use these generalized boundaries
for US states and US counties. New projects may entail the production of map cari-
catures or generalizations when the available boundaries are not suitable.
he third purpose of micromaps is to show geospatial patterns. Figure
.
shows
individual perceptualgroups ofstates. hissimple version isgoodas an introduction
and reveals some geospatial grouping. More sophisticated variations on micromaps
have also called attention to contours constructed from region polygons. One ap-
proach shows background states in light grey with white outlines. States highlighted
previously in a perceptual group can appear in an additional color such as light yel-
lowandhaveablackoutline.henongreyblack-outlined states thenappearasafore-
ground contour. Examples with two complementary contours for regions above and
below the median will be further discussed in Sects.
.
.
and
.
.
.
Web-based Applications of LM Plots
1.4
Overthelastdecade,USFederalAgenciesandotherinstitutionshaveincreasinglyfo-
cusedattention ondistributing large amounts of geographically referenced statistical
data, either inprint orthroughthe Web.heWeb-baseddistribution ofdata isaimed
at replacing printed tabular displays and at providing access to current data quickly.
Several approaches have been developed that provide a user-friendly Web-based in-
terface to tabular and graphical displays of federal data. he user can interactively
and dynamically query and sort the data, compare different geographic regions, and
look at the data at different spatial resolutions, e.g., at the state or the county level.
