Movatterモバイル変換

This package provides tools for exploring and describing longitudinalnetwork data. It works with temporal network data that is stored as anetworkDynamic object – essentially a list in thenetwork format in which elements also have an attachedactivity attribute, a matrix of spells indicating when vertex or edge isactive. See thenetworkDynamicpackage for more information on translating longitudinal network datafrom other formats (timed edge lists, lists of toggles, sets ofmatrices, etc).

Thetsna package contains functions for calculating arange of descriptive statistics from longitudinal network data. Some ofthese are simple temporal extensions of static network descriptors,others are unique to longitudinal network data.

• Timeseries of standard ‘static’ social network analysismetrics

  These functions operate by collapsing the dynamic network into astatic network at a series of regular intervals and returning theresults as a time series ts object. They can provide general descriptionof trends in a network dataset.

  • tSnaStats – descriptive statistics from thesna package. Includesboth graph- and vertex-level measures such as centralities, components,reciprocity, betweenness, triad-census, etc.
  • tErgmStats – descriptive statistics for the “terms”used in statistical Exponential-family Random Graph Models (ERGMs) fromtheergmpackage

• Temporal path based metrics

  These functions compute and use temporal paths (network geodesicsthat are constrained by the activity times of edges) through anetwork.

  • tPath
  • tReach

• Rates and Duration metrics

  These functions can be used to compute distributions of (observed)activity durations in a data structure. Note that due to censoring(edges that begin before or end after the time observation window forthe network) the observed durations may be biased away from the ‘real’values (or model parameters). The duration estimate functions usevarious types of survival analysis to return estimates of thesevalues.

  • edgeDuration
  • vertexDuration
  • tiedDuration
  • tEdgeDissolution
  • tEdgeFormation
  • tiedDuration

• Utility functions

  These provide tools for data manipulation and plotting

  • as.network.tPath
  • plotPaths
  • timeProjectedNetwork

Thetsna package is part of thestatnet software suite for networkanalysis written inR, and it is designed to workseamlessly with the otherpackages in this suite.

The development of this software was supported by grant R01HD68395from the National Institute of Health.

Docs and examples

Thepackagevignette (browseVignettes(package='tsna')) providesexamples and illustrations of key concepts.

Examples of howtsna tools might be used in a networkdata analysis workflow can be found in ourWorkshop materials, inparticular theTemporalnetwork tools andtergm workshops.

License and attribution

This software is distributed under the GPL-3 license. It is free,open source, and has the attribution requirements (GPL Section 7) athttps://statnet.org/attribution

To cite package ‘tsna’ in publications please use: Bender-deMoll S,Morris M (2021).tsna: Tools for Temporal Social NetworkAnalysis. R package version 0.3.5,https://CRAN.R-project.org/package=tsna.

Code of Conduct

Developers and Contributing Users to the Statnet Project should readhttps://statnet.github.io/private/ for information about therelationship between the public and the private repository and theworkflows involved.

Please note that the tsna project is released with aContributorCode of Conduct. By contributing to this project, you agree to abideby its terms.