r packages
Why develop R packages?
Many of my research projects also involve developing R packages. But, I'm not a software developer, so why bother? These R packages are designed to help make research easier. Each one includes detailed documentation and a complete vignette illustrating how to use it. They are also designed following the tinyverse philosophy, which helps ensure they will keep working in the future. This page provides brief descriptions of each package, but you can find more information about the associated projects on the current and past project pages.
​
Each package has its own Hex logo. If you'd like a sticker, let me know.
backbone
The backbone package provides methods to identify the most significant links in a network and to extract its backbone. For example, given a weighted network, returns a new network that only contains the edges that are statistically significantly strong (i.e., the backbone). It offers methods for weighted networks, bipartite projections, and dense unweighted networks. It is available from CRAN and can be installed by typing install.packages("backbone") into the R console. You can find a detailed vignette here.
incidentally
The incidentally package provides methods for generating incidence matrices that represent bipartite networks or hypergraphs. These matrices can be generated so that they have specific characteristics, or can be generated from a unipartite network using one of a series of generative models. The package also provides an API for downloading and generating an incidence matrix or bipartite network representing bill sponsorship in the US Congress that can be customized by session, chamber, and topic. It is available from CRAN and can be installed by typing install.packages("incidentally") into the R console. You can find a detailed vignette here.
childfree
The childfree package imports demographic data from a variety of public data sources, extracting and harmonizing variables that are useful for studying childfree individuals. Currently it provides access to data from the MSU State of the State Survey (Michigan), the CDC National Survey of Family Growth (United States), and the USAID Demographic and Health Surveys Program (multiple developing countries). It is available from CRAN and can be installed by typing install.packages("childfree") into the R console. You can find a detailed vignette here.
KOLaide
The KOLaide package assists in the selection of Key Opinion Leaders (KOLs) who can facilitate diffusion in a network. Unlike keyplayer packages that identify the one optimal set of nodes, KOLaide is designed to address practical challenges that often arise in implementation settings. For example, users can specify which network members are available to serve as KOLs, a range of KOL teams to consider, and characteristics on which KOL teams should be diverse. It is available from CRAN and can be installed by typing install.packages("KOLaide") into the R console. A vignette is coming soon, but it is described in detail in this paper.
grand
The grand package is designed to facilitate using the Guidelines for Reporting About Network Data (GRAND). Currently is has limited functionality, and serves as a placeholder that will be expanded when the final guidelines are developed. It is available from CRAN and can be installed by typing install.packages("grand") into the R console. You can find a detailed vignette here.