spsann (version 1.0.1)

spsann-package: spsann: Optimization of Sample Configurations using Spatial Simulated Annealing

Description

Methods to optimize sample configurations using spatial simulated annealing. Multiple objective functions are implemented for various purposes, such as variogram estimation, trend estimation, and spatial interpolation. A general purpose spatial simulated annealing function enables the user to define his/her own objective function.

Support

spsann was initially developed as part of the PhD research project entitled Contribution to the Construction of Models for Predicting Soil Properties, developed by Alessandro Samuel-Rosa under the supervision of Lúcia Helena Cunha dos Anjos (Universidade Federal Rural do Rio de Janeiro, Brazil), Gustavo de Mattos Vasques (Embrapa Solos, Brazil), and Gerard B. M. Heuvelink (ISRIC - World Soil Information, the Netherlands). The project was/is supported from 2012 to 2016 by the CAPES Foundation, Ministry of Education of Brazil, and the CNPq Foundation, Ministry of Science and Technology of Brazil.

Contributors

Some of the solutions used to build spsann were found in the source code of other R-packages. The skeleton of the optimization functions was adopted from the intamapInteractive package, authored by Edzer Pebesma [e-mail] and Jon Skoien [e-mail].

A few small solutions were adopted from the SpatialTools package, authored by Joshua French [e-mail], and clhs package, authored by Pierre Roudier [e-mail].

Conceptual contributions were made by Gerard Heuvelink [e-mail], Dick Brus [e-mail], Murray Lark [e-mail], and Edzer Pebesma [e-mail].

Details

spsann is the R package for the optimization of sample configurations using spatial simulated annealing. It includes multiple functions with different objective functions to optimize sample configurations for variogram estimation (number of points or point-pairs per lag distance class), trend estimation (association/correlation and marginal distribution of the covariates), and spatial interpolation (mean squared shortest distance). spsann also includes objective functions that can be used when the model of spatial variation is known (mean (or maximum) kriging variance). Some objective functions were combined to optimize sample configurations when we are ignorant about the model of spatial variation (terra incognita). A general purpose function enables to user to define his/her own objective function and plug it in the spatial simulated annealing algorithm.

Spatial simulated annealing is a well known method with widespread use to solve optimization problems in the environmental sciences. This is mainly due to its robustness against local optima. At each iteration, the algorithm evaluates if a worsening solution can be accepted. The chance of accepting worsening solutions reduces as the number of iterations increases.

ll{ Package: spsann Type: Package Version: 1.0.1 Date: 2015-07-14 License: GPL (>= 2) }