README.md

In jarioksa/natto: An Extreme 'vegan' Package of Experimental Code

Welcome to natto!

natto is an extreme vegan package that contains experimental, specialized or unfinished code that does not fit to the more general vegan package. Some of these functions may be merged to vegan later, but the package is not a code kindergarten with guaranteed tenure track to vegan.

Currently natto contains

• Rao's quadratic entropy, phylogenetic dissimilarities and phylogenetic data standardization, plus related Clarke's taxonomic dissimilarity.
• Polar Ordination A.K.A. Bray-Curtis Ordination.
• Orlóci's Position Vector Ordination and a related method using species as ordination axes.
• Implementation of Decorana in R: easier to study the algorithm and develop the method than with Fortran code.
• Estimate residual species associations in CCA and RDA.
• Randomized clustering and ordination based on Bayesian Jaccard dissimilarity with applications. Dissimilarities are random numbers from Beta distribution and each evaluation yields a new set of dissimilarities. For clustering, this can be used to assess the support of clusters, and for ordination the confidence areas of coordinates.
• Hierarchical clustering based on information analysis after Williams, Lambert & Lance, J. Ecol. 54, 427-445 (1966).
• beta and gamma diversity clustering.
• Plotting cluster dendrograms (hclust results) so that terminal leaves are fans with base proportional to weight (size): ginkgogram.
• Humped no-interaction model for species richness vs. biomass (transferred from the vegan package).
• IAP functions to find the indicator values for species and statistics for sites in Index of Atmospheric Purity. The main focus is in indicator values for species which are the number of companion species. These can be used to find species that indicate high or low biodiversity, and permutation test is provided to assess these values.
• Mean Rank Shift function of Collins et al., Ecology 89, 3534-3541 (2008).
• Testing differences of slopes in Mantel style regression between dissimilarities.
• Casting unconstrained vegan::rda() result object to stats::prcomp object.
• Casting Delaunay triangulation from deldir::deldir result object to stats::dist object.

Installation

There are no binary packages available at the moment, but if you install Hadley Wickham's devtools package from CRAN, you can install natto directly from github:

install.packages("devtools") # if you have not yet installed devtools
library(devtools)
install_github("jarioksa/natto")

Binary package can be installed via R-Universe:

install.packages("natto", repos = c("https://jarioksa.r-universe.dev",
"https://cloud.r-project.org"))

jarioksa/natto documentation built on Aug. 30, 2024, 10:18 a.m.