The hardware and bandwidth for this mirror is donated by METANET, the Webhosting and Full Service-Cloud Provider.
If you wish to report a bug, or if you are interested in having us mirror your free-software or open-source project, please feel free to contact us at mirror[@]metanet.ch.

mpaR 0.5.0

Performance

• Rewrote the core dynamic-programming routines to be vectorised and layered. traversal_weights(), node_weights(), and all three main_path() types (global, local, key-route) now run one vectorised step per topological layer instead of one interpreted igraph call per vertex. On a 50,000-vertex / 300,000-arc citation network the key-route extraction drops from tens of minutes to a few seconds, with identical results (verified against Pajek and by exhaustive path enumeration on small graphs).
• read_pajek() now parses arcs in a single vectorised pass. The previous version built one data frame per arc and was effectively O(E^2), taking minutes on large networks; it is now sub-second.

New features

• key_route_sweep(): computes the key-route main path for many values of k in a single shared DP pass, returning a tidy k / n_nodes / n_arcs table (and optionally the subgraphs). This is the efficient way to choose k by looking for a plateau, replacing a slow loop over main_path().
• main_path(type = "key_route") gains a k_range = c(start, end) argument to seed from a slice of key routes ranked start to end by weight, mirroring Pajek’s key-route range option.

mpaR 0.4.0

Initial CRAN release.

Features

• Implements Main Path Analysis (MPA) as introduced by Hummon and Doreian (1989), computing traversal weights (SPC, SPLC, SPNP) for each edge of a directed acyclic graph (DAG) and extracting the global, local, and key-route main paths.
• check_dag() validates that a graph is acyclic and reports the back-edges responsible for any cycles.
• classify_nodes() labels vertices by role (source/terminal/user, or source/sink/user under the graph-theory convention).
• check_scale_free() fits a discrete power-law to the degree distribution and tests for scale-free behaviour.
• edge_weights() and node_weights() compute traversal weights at the edge and node level.
• Readers for Pajek (read_pajek()) and Gephi export (read_gephi_export(), supporting .gexf and .graphml) files.
• plot_mpa() visualizes main paths, with a scope argument for restricting the plot to a subgraph.
• Per-component path extraction for disconnected networks.
• Accepts igraph objects or edge-list data frames as input throughout.

These binaries (installable software) and packages are in development.
They may not be fully stable and should be used with caution. We make no claims about them.