Geoscience Reference
In-Depth Information
shown in the
mappingtools
code chunk in Appendix 17B. This demonstrates not only the use of
spatial data but also that the system is capable of including graphical output from the incorporated
code. As before, the R commands in the
Rnw
document are echoed to the final
pdf
document to
illustrate the method, although in more usual situations this would not be the case.
This example also illustrates the incorporation of detail giving the source of the data. Here,
the income data itself is obtained via a website provided by the Centers for Disease Control and
Prevention (CDC)* and the map boundary data is obtained via a website provided by the US Census
Bureau.
†
The embedded R code contains instructions to download, unzip and read these files into
R - the former is in
dbf
format, the latter in ESRI shapefile format - the code also contains com-
ments showing the URL of human-readable web pages describing the data (these links are also
given in the footnotes referred to earlier). This code is listed in the chunk labelled
dataprep
in
Appendix 17B.
17.3.3 gc e
xaMPle
: P
ycnoPhylactic
i
nterPolation
As a further example, Figure 17.2 shows the result of applying
pycnophylactic interpolation
( Toble r,
1979) to another variable from the 1990 US Census; in this case, the variable under scrutiny is the
number of children aged 1 year and under, on a per-county basis. This variable is obtained from
the same source as the median household income data used in the last example. This is a technique
used to estimate population density as a continuous surface, given a list of population counts for
a set of supplied corresponding geographic zones. The population counts in this example are the
Pycnophylactic surface of population under 1 year old
(Georgia counties)
FIGURE 17. 2
Incorporated R code producing graphical output - pycnophylactic surface. (From 1990 US
Census.)
*
http://www2.cdc.gov/nceh/lead/census90/house11/houintro.htm.
†
http://www.census.gov/geo/www/cob/co1990.html.
