bafuentes / rassta

R package for landscape stratification, sampling, and modeling of environmental phenomena using spatial data in R.

rassta

rassta (raster-based spatial stratification algorithms) is a collection of algorithms for the spatial stratification of landscapes, sampling, and modeling of spatially-varying phenomena in R.

rassta offers a simple framework for the stratification of geographic space based on raster layers representing landscape factors and/or factor scales. The stratification process follows a hierarchical approach, which is based on first level units (i.e., classification units) and second-level units (i.e., stratification units). Nonparametric techniques allow to measure the correspondence between the geographic space and the landscape configuration represented by the units. These correspondence metrics are useful to define sampling schemes and to model the spatial variability of environmental phenomena.

Installation

rassta is available from CRAN, so the current released version can be installed as follows:

install.packages("rassta")

To install the development version from github, please use:

remotes::install_github("bafuentes/rassta")

Cheat Sheet

Full documentation of rassta including some vignettes can be found here

Acknowledgments

rassta greatly benefits from past and current efforts to make spatial data analysis fully operational in R, which in turn have benefited from titans like GDAL, PROJ, GEOS, etc. Special thanks to the minds behind the terra, rgdal, rgeos, and sf packages, and those behind the packages that rassta depends on as well.

Citation

A journal article describing the methods and theoretical background of rassta has been accepted by the R Journal and now it is in press. In the meantime, if you use rassta as part of your research please cite the EarthArXiv preprint in your work:

B.A. Fuentes, M.J. Dorantes, and J.R. Tipton. rassta: Raster-Based Spatial Stratification Algorithms. EarthArXiv, 2021. https://doi.org/10.31223/X50S57

rassta's Issues

Error message: "Error: [] lengths of of cells and values do not match"

I am having an issue running the som_pam function where I a get the Error: [] lengths of of cells and values do not match message.
Upon examining the code for som_pam and running the code line by line I found out that the issue occurs at r.som[idx] <- terr.som$SOM$unit.classif. Essentially, for some reason the number of values don't match, i.e. length(values(r.som)) == length(terr.som$SOM$unit.classif) is FALSE.

Can you think of a reason why this may be occurring? Are there parameters I should alter in the som_gap function as I assume something is happening here where the number of cells in the input rasters is being altered when producing the classification units.