A Suite of Packages for the Statistical Modeling of Network Data

Share:

Description

statnet is a collection of software packages for statistical network analysis that are designed to work together, and provide seamless access to a broad range of network analytic and graphical methodology. The packages implement recent advances in network modeling based on exponential-family random graph models (ERGM), as well as latent space models and more traditional network methods. Together, the packages provide a comprehensive framework for ERGM-based cross-sectional and dynamic network modeling: tools for model estimation, model evaluation, model-based network simulation, and network visualization. The statistical estimation and simulation functions are based on a central Markov chain Monte Carlo (MCMC) algorithm. The coding is optimized for speed and robustness.

The code is actively developed and maintained by the statnet development team. New functionality is being added over time.

Details

statnet packages are written in a combination of R and C It is usually used interactively from within the R graphical user interface via a command line. it can also be used in non-interactive (or “batch”) mode to allow longer or multiple tasks to be processed without user interaction. The suite of packages are available on the Comprehensive R Archive Network (CRAN) at http://www.r-project.org/ and also on the statnet project website at http://statnet.org/

The statnet suite of packages has the following components:

For data handling:

  • network is a package to create, store, modify and plot the data in network objects. The network object class, defined in the network package, can represent a range of relational data types and it supports arbitrary vertex / edge /graph attributes. Data stored as network objects can then be analyzed using all of the component packages in the statnet suite. (automatically downloaded)

  • networkDynamic extends network with functionality to store information about about evolution of a network over time, defining a networkDynamic object class. (automatically downloaded)

For analyzing cross-sectional networks:

  • ergm is a collection of functions to fit, simulate from, plot and evaluate exponential random graph models. The main functions within the ergm package are ergm, a function to fit linear exponential random graph models in which the probability of a graph is dependent upon a vector of graph statistics specified by the user; simulate, a function to simulate random graphs using an ERGM; and gof, a function to evaluate the goodness of fit of an ERGM to the data. ergm contains many other functions as well. (automatically downloaded)

  • ergm.count is an extension to ergm enabling it to fit models for networks whose relations are counts. (automatically downloaded)

  • latentnet is a package to fit and evaluate latent position and cluster models for statistical networks The probability of a tie is expressed as a function of distances between these nodes in a latent space as well as functions of observed dyadic level covariates. (optional download)

  • sna is a set of tools for traditional social network analysis. (automatically downloaded)

  • degreenet is a package for the statistical modeling of degree distributions of networks. It includes power-law models such as the Yule and Waring, as well as a range of alternative models that have been proposed in the literature. (optional download)

For temporal (dynamic) network analysis:

  • tergm is a collection of extentions to ergm enabling it to fit discrete time models for temporal (dynamic) networks. The main function in tergm is stergm (the “s” stands for separable), which allows the user to specify one ergm for tie formation, and another ergm for tie dissolution. The models can be fit to network panel data, or to a single cross-sectional network with ancillary data on tie duration. (automatically downloaded)

  • tsna is a collection of extensions to sna that provide descriptive summary statistics for temporal networks. (optional download)

  • relevent is a package providing tools to fit relational event models. (optional download)

Additional utilities:

  • ergm.userterms provides a template for users who want to implement their own new ERGM terms. (separate download required)

  • networksis is a package to simulate bipartite graphs with fixed marginals through sequential importance sampling. (optional download)

  • EpiModel is a package for simulating epidemics (optional download)

statnet is a metapackage; its only purpose is to provide a convenient way for a user to load all of the packages in the statnet suite. It does this by depending on all of the packages, so that loading the statnet package into R automatically loads all packages above that are labeled "automatically downloaded". If the user specifies install.packages("statnet",dependencies=T), statnet will also download all of the packages above that are labeled "optional download". Those can, of course, also be installed individually.

Each package in statnet has associated help files and internal documentation, and additional the information can be found on the Statnet Project website (http://statnet.org/). Tutorials, instructions on how to join the statnet help mailing list, references and links to further resources are provided there. For the reference paper(s) that provide information on the theory and methodology behind each specific package use the citation("packagename") function in R after loading statnet.

We have invested much time and effort in creating the statnet suite of packages and supporting material so that others can use and build on these tools. All we ask in return is that you cite it when you use it. For publication of results obtained from statnet, the original authors are to be cited as described in citation("statnet"). If you are only using specific package(s) from the suite, please cite the specific package(s) as described in the appropriate citation("packgename"). Thank you!

Author(s)

Mark S. Handcock handcock@stat.ucla.edu,
David R. Hunter dhunter@stat.psu.edu,
Carter T. Butts buttsc@uci.edu,
Steven M. Goodreau goodreau@uw.edu,
Pavel N. Krivitsky pavel@uow.edu.au, Skye Bender-deMoll skyebend@skyeome.net and
Samuel Jenness (for EpiModel) sjenness@uw.edu Martina Morris morrism@uw.edu

Maintainer: Martina Morris morris@uw.edu

Want to suggest features or report bugs for rdrr.io? Use the GitHub issue tracker.