Please make sure you are using the code and docs maintained by the Smithsonian Data Science Lab (github.com/sidataciencelab/ecostructure).
This package contains functions for fitting STRUCTURE type models to ecological data, both at local and global geographic scales, together with GIS based visualizations of the fitted models. These grade-of-membership models can be used to assess the local representation of large regional biotas, their degree of intermixing in local assemblages, and their rate of turnover across geographic space owing to environmental or climatic turnover. ecostructure makes use of advances in clustering algorithms, first from the package CountClust in 0.99.1 and now leveraging its successor fastTopics in 2.0.
Install ecostructure following the instructions below.
remotes::install_github("stephenslab/fastTopics")
remotes::install_github("linxihui/NNLM")
remotes::install_github("sidatasciencelab/ecostructure")
ecostructure requires access to the "gfortran" library. Mac OS X users may encounter the error "library not found for -lgfortran" when installing. To fix this error, please follow the instructions at this link. Apple Silicon users should instead follow the directions found here.
Then load ecostructure
library(ecostructure)
Some examples of visualizations produced using our ecostructure package
If you want to try ecostructure and replicate figures like this, please check our tutorial here.
If you are using ecostructure or our code, please cite our papers:
White, Alexander E. and Dey, Kushal K. and Mohan, Dhananjai and Stephens, Matthew and Price, Trevor D. Regional influences on community structure across the tropical-temperate divide. Nature Communications. 2019. 10 (1). 2646. 10.1038/s41467-019-10253-6
White, Alexander E. and Dey, Kushal K. and and Stephens, Matthew and Price, Trevor D. Dispersal syndromes drive the formation of biogeographical regions, illustrated by the case of Wallace’s Line. Global Ecology and Biogeography. 2021. 10.1111/geb.13250
For any queries or concerns related to the software, you can open an issue here.
The methods used to develop this framework have advanced since our initial release and may yet evolve in future iterations of this package. The Smithsonian Data Science Lab continues to collaborate with the Dey Lab (Sloan Kettering), the Price Lab (U Chicago) and the Stephens Lab (U Chicago) to develop this work.
You are welcome to contribute to ecostructure by forking this repo.