|
|
|||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| Pfad: | www.math.uni-klu.ac.at : / stat / users / agebhard / ERSA-98-D8-427 |
|
|
Next: Introduction Implementing functions for spatial statistical analysis using the R language1Roger Bivand
August 1998
Abstract:R is a language similar to S for statistical data analysis,
based on modern programming concepts and released under the GNU
General Public License. It permits the integration of program
scripts with compiled dynamically loaded libraries of functions when
computing speed is important. Following a broad outline of existing
collections of functions for spatial statistics written for S , we
show how they may be ported to R , and compare their
characteristics. We further demonstrate how existing work may be
extended to topics not yet covered, and how libraries of functions
may be constructed to give the analyst insight into the data sets
under analysis.
Functions for three types of spatial statistics are covered:
spatially continuous data, point pattern data, and area data.
Geographical information systems are a major source of such data,
and since access to source code of R permits the evaluation of
alternative methods for linking it with GIS, we discuss how such
links impact the implementation of analytical functions. In
particular we examine the specification of data frames for spatial
data, and which classes and methods provided in R are appropriate
for handling spatial data.
We conclude by presenting packaged R functions for spatial
statistical analysis, and their application to standard data sets
(included in the distributed software). Both the development of R ,
and of these functions, is on-going, but seem to us to have reached
a critical mass making R an attractive platform for doing spatial
statistics.
Next: Introduction Albrecht Gebhardt 1998-08-29 |
|
|||||
|
