SSN: AnRPackage for Spatial Statistical Modeling on Stream Networks

• DOI: 10.18637/jss.v056.i03
• Publication Date: 2015-09-21T20:30:55Z
• Authors: Jay M. Ver Hoef; Erin E. Peterson; David Clifford; Rohan Shah
• Keywords: Data set

The SSN package for R provides a set of functions modeling stream network data. can import geographic information systems data or simulate new as 'SpatialStreamNetwork', object class that builds on the spatial sp classes. Functions are provided fit linear models (SLMs) 'SpatialStreamNetwork' object. covariance matrix SLMs use distance metrics and geostatistical unique to networks; these account distances topological configuration networks, including volume direction flowing water. In addition, traditional Euclidean simple random effects included, along with Poisson binomial families, generalized mixed model framework. Plotting diagnostic provided. Prediction (kriging) be performed missing separate unobserved locations, block prediction (block kriging) used over sets segments. This article summarizes importing, simulating, data, diagnostics prediction.