The most commonly used R-packages for estimating psychological networks at present are listed below. Many of these are under development, and it is advised to make sure to always install the most recent versions of the packages from the authors’ Github pages via the devtools package.

Cross-sectional data
  • qgraph | For estimating so-called Gaussian Graphical Models in cross-sectional data – networks that consist of ordinal or Gaussian variables. qgraph is also a great package to visualize all kinds of different networks.
  • IsingFit | For estimating so-called Ising Models in cross-sectional data (networks with binary variables).
  • mgm | For estimating Mixed Graphical Models in cross-sectional data – network models that consist of different types of variables.
  • bootnet | For estimating the accuracy and stability of cross-sectional network models.
  • NetworkComparisonTest | For testing the difference of network structures across different samples, and across the same sample over time.
Time-series data
  • graphicalVAR | For estimating time-series network models (vector autoregression models; VAR) in case of n=1.
  • mlVAR | For estimating time-series network models (multilevel vector autoregression models; multilevel VAR) in case of n>1.
  • mgm | For estimating time-varying time-series networks models (i.e. network models that can change over time).