- Added a
NEWS.md file to track changes to the
package.
- The bridge() function is now available. Can be used to calculate and
plot “bridge statistics”, including bridge strength, bridge betweenness,
bridge closeness, and bridge expected influence.
- Bug fixes for bridge strength in the bridge() function
- The goldbricker() and net_reduce() functions are now available.
Useful for comparing correlations in order to eliminate nodes in
networks which ostensibly measure the same construct
- Additions to bridge() function: can use list input for communities,
improved error handling
- New argument “corMin” in goldbricker function
- Three new functions for plotting networks using MDS, PCA, and
eigenmodels (MDSnet, PCAnet, EIGENnet)
- Added an argument in bridge() to average bridge centrality by number
of nodes in communities
- MDSnet, EIGENnet, & PCAnet now include an optional repulsion
parameter. This allows the user to avoid any node overlap in plots.
Additional arguments can now be passed to qgraph via …
- The “average” argument in bridge has been changed to “normalize”
& normalizes bridge centralities
- Added PROCRUSTESnet function for plotting two networks with
Procrustes-aligned MDS
- In plot.bridge, argument plotNA (default FALSE) now controls whether
nodes with NA values are plotted
- Removed previously deprecated functions edge.impact, global.impact,
and structure.impact
- Fixed a bug in impact.NCT where multiple booleans were passed to if
statement
- Fixed deprecated /donttest examples
- Fixed a bug in bridge that resulted in incorrect communities if a
non-sequential list was supplied, e.g., communities <-
list(“Comm1”=c(1:2, 4), “Comm2”=c(3))
- Amended the previous fix to work in a wider variety of cases
- Adjusted goldbricker to handle tibbles
- Added testing with testthat
- Removed deprecated impact functions, resolving circular dependency
on NetworkComparisonTest package
- Update package description to conform to roxygen2 7.0.0
- Fixed a bug affecting the zou2007 method of goldbricker
- Amends an error in the calculation of bridge closeness introduced in
1.4.0. Versions 1.4.0-1.5.2 mistakenly used the mean of inverse
distances (rather than the inverse of the mean distance) for networks
without negative edges