Transition Network Analysis (TNA)

Provides tools for performing Transition Network Analysis (TNA) to study relational dynamics, including functions for building and plotting TNA models, calculating centrality measures, and identifying dominant events and patterns. TNA statistical techniques (e.g., bootstrapping and permutation tests) ensure the reliability of observed insights and confirm that identified dynamics are meaningful. See (Saqr et al., 2025) for more details on TNA.

Additional Layout Algorithms for Network Visualizations

Several new layout algorithms to visualize networks are provided which are not part of 'igraph'. Most are based on the concept of stress majorization by Gansner et al. (2004) . Some more specific algorithms allow the user to emphasize hidden group structures in networks or focus on specific nodes.

Similarity Network Fusion

Similarity Network Fusion takes multiple views of a network and fuses them together to construct an overall status matrix. The input to our algorithm can be feature vectors, pairwise distances, or pairwise similarities. The learned status matrix can then be used for retrieval, clustering, and classification.

Siena - Simulation Investigation for Empirical Network Analysis

The main purpose of this package is to perform simulation-based estimation of stochastic actor-oriented models for longitudinal network data collected as panel data. Dependent variables can be single or multivariate networks, which can be directed, non-directed, or two-mode; and associated actor variables. There are also functions for testing parameters and checking goodness of fit. An overview of these models is given in Snijders (2017), .

Bayesian Regularization for Feed-Forward Neural Networks

Bayesian regularization for feed-forward neural networks.

Bootstrap Methods for Various Network Estimation Routines

Bootstrap methods to assess accuracy and stability of estimated network structures and centrality indices . Allows for flexible specification of any undirected network estimation procedure in R, and offers default sets for various estimation routines.

Latent Position and Cluster Models for Statistical Networks

Fit and simulate latent position and cluster models for statistical networks. See Krivitsky and Handcock (2008) and Krivitsky, Handcock, Raftery, and Hoff (2009) .

R Interface to 'Keras'

Interface to 'Keras' < https://keras.io>, a high-level neural networks 'API'. 'Keras' was developed with a focus on enabling fast experimentation, supports both convolution based networks and recurrent networks (as well as combinations of the two), and runs seamlessly on both 'CPU' and 'GPU' devices.

An Simplified Implementation of the 'network' Package Functionality

An implementation of some of the core 'network' package functionality based on a simplified data structure that is faster in many research applications. This package is designed for back-end use in the 'statnet' family of packages, including 'EpiModel'. Support is provided for binary and weighted, directed and undirected, bipartite and unipartite networks; no current support for multigraphs, hypergraphs, or loops.

Data Analysis for IP Addresses and Networks

Classes and functions for working with IP (Internet Protocol) addresses and networks, inspired by the Python 'ipaddress' module. Offers full support for both IPv4 and IPv6 (Internet Protocol versions 4 and 6) address spaces. It is specifically designed to work well with the 'tidyverse'.