I’ve written a few R packages to support my work in co-occurrence analysis, text analysis, and visualising complex datasets. You can find some basic descriptions below, while the packages themselves are available on GitHub. The simplest way to download them in R is via the ‘devtools’ package, as follows:
library(devtools) # This is available on CRAN install_github('mjwestgate/revtools') # Not on CRAN library(revtools) # load
revtools: Tools to support systematic review in R
This is a new package that imports bibliographic data from sources like Endnote, Web of Science or Scopus, and uses topic modelling to visualise patterns in that data. It is built in Shiny, so users can interact with the data in a range of interesting ways, and can select and export key references. This is in early development, so any feedback is welcome.
circleplot: plots of distance and association matrices
I made this code to draw simple, pretty plots of association and correlation networks between species, individuals or concepts. It was inspired by the circlize package, though my code is much simpler and with far fewer options. My rationale was that using curves rather than straight lines, and careful use of colours, can improve the interpretability of plots of this kind of data.
sppairs: Species Pairwise Association Analysis (SPAA)
SPAA is a method for investigating species co-occurence patterns using odds ratios. sppairs was originally built to implement this method in the R language; but later versions converted this to a generic apply function for calculation of a range of pairwise statistics. This package is still available on GitHub, but is no longer under active development.