This function combines segmentation of scattered 2D data and estimation of anisotropy parameters using the CTI method.
anisotropyChoice(object)object: A modified Intamap type object is returned,
which contains the results of the anisotropy parameter estimation.
The anisotropy parameters are returned in the element anisPar
as described below.
List element in object that contains a list with the following
elements:
ratioA coarse-grained anisotropy ratio for all the data
directionA coarse-grained anisotropy orientation for all the data
clustersA matrix of dimension [cl]x5 which determines the anisotropy per cluster.
Each row of clusters gives the
(cluster id, anisotropy ratio, anisotropy direction, number of points, area) for each cluster detected.
list element added to the original object containing the segmentation results.
Index array identifying the cluster in which each observation point belongs. Zero value means that the observations has been removed.
Number of clusters detected.
An Intamap type object containing one
SpatialPointsDataFrame with observations.
D.T. Hristopulos, G.Spiliopoulos, A.Chorti
The function AnisotropyChoice function employs the
doSegmentation function to
automatically separate the original dataset into clusters based on the sampling density and the spatial
locations of the data (see doSegmentation for
details). The results of the segmentation procedure and the
anisotropy analysis per cluster are returned in a matrix of
dimension [cl]x5, where [cl] is the number of clusters . Each row of
the matrix contains the cluster index, the anisotropy ratio, the
anisotropy direction, the number of cluster points and the area
inside the convex hull of the cluster. In addition, a single set of
anisotropy parameters is returned in the element anisPar.
These parameters are calculated using weighted averages of the
covariance Hessian matrix estimates in each cluster. The weights are
based on the area enclosed by the convex hull of each cluster.
[1] http://www.intamap.org
[2] A. Chorti and D. T. Hristopulos (2008). Non-parametric Identification of Anisotropic (Elliptic) Correlations in Spatially Distributed Data Sets, IEEE Transactions on Signal Processing, 56(10), 4738-4751 (2008).
[3] D. T. Hristopulos, M. P. Petrakis, G. Spiliopoulos, A. Chorti (2009). Non-parametric estimation of geometric anisotropy from environmental sensor network measurements, StatGIS 2009: Geoinformatics for Environmental Surveillance Proceedings (ed. G. Dubois).
library(gstat)
data(walker)
object=createIntamapObject(observations=walker)
object=anisotropyChoice(object)
print(summary(object$clusters$index))
print(object$anisPar)
Run the code above in your browser using DataLab