Releases: szhorvat/IGraphM
Releases · szhorvat/IGraphM
v0.6.5
v0.6.4
Added:
IGBetaWeightedGabrielGraph
for computing the β value where each edge of a lune-based β-skeleton would disappear (experimental function).IGModularityMatrix
gives the modularity matrix of a network.IGCanonicalLabeledGraph
creates a canonical representation of a labeled graph without changing vertex names. Graphs which are considered equivalent byIGSameGraphQ
have the same canonical representation. It is intended for removing duplicate graphs usingDeleteDuplicatesBy
.IGCanonicalEdgeList
creates a canonical representation of an edge list, in a manner similar toIGCanonicalLabeledGraph
.
Other:
- Documentation improvements
- Significant performance improvements for the calculation of lune and circle based β skeletons when using large β values.
v0.6.3
This is a maintenance release, recommended for all users.
0.6.3
Changed:
- When an invalid vertex is passed to IGraph/M functions, now the message shown is
IGraphM::invv
instead ofVertexIndex::inv
.
Fixed:
- In rare cases, vertex names that were lists of other vertex names in the same graph were misinterpreted. This affected only Mathematica 12.0 and later.
IGCommunitiesLeadingEigenvector
is less likely to fail to converge now.IGCommunitiesLeadingEigenvector
now respects random seeds set in Mathematica. These affect only the random starting vector it uses internally, which can very rarely cause non-deterministic output.IGBarabasiAlbertGame
no longer accepts negative or zero exponents, as these were not handled correctly.IGBridges
is no longer prone to stack overflow when given very large graphs.
Other:
- Performance improvements for
IGGraphEditor
. - Documentation improvements.
v0.6.2
0.6.2
Changes:
- Re-enabled resizing for
IGGraphEditor[]
. IGGraphEditor[]
now shows a warning when invoked without a notebook interface.
Fixed:
IGCommunitiesSpinGlass
returned invalid results for the null and singleton graphs.
Other:
- Documentation improvements.
v0.6.1
0.6.1
Fixed:
IGCompareCommunities
had a corrupted usage message.- Restored compatibility with older Linux distros that use GCC 4.8, such as RHEL / CentOS 7.
- The Fruchterman-Reingold, Kamada-Kawai and DrL layouts now verify that all weights are positive.
IGAsymmetrixPreferenceGame
misinterpreted the type weight matrix.
Other:
- Temporarily disabled resizing in graph editor, as it interferes with vertex creation.
- Documentation improvements.
v0.6.0
0.6.0
This version is based on the 0.9 series of C/igraph.
Added:
IGHarmonicCentrality
andIGHarmonicCentralityCutoff
compute the harmonic centrality and range-limited harmonic centrality.IGLinkRank
andIGPersonalizedLinkRank
compute the equivalent of PageRank for edges.IGNeighborhoodCloseness
computes the range-limited closeness centrality, as well as the number of vertices reachable within the given range.IGFamousGraph
exposes the igraph C library's built-in graph database.IGPreferenceGame
andIGAsymmetricPreferenceGame
create non-growing random graphs based on vertex types.IGReingoldTilford
andIGReingoldTilfordCircular
now support theDirectedEdges
option.IGFruchtermanReingold
now supports constraining the coordinates of a subset of vertices.IGPercolationCurve
for efficiently computing the size of the largest component as a function of mean degree while removing edges.IGShortestPathTree
for computing a shortest path tree rooted in a given vertex.IGGraphEditor
is an experimental interactive graph editor.- Experimental progress reporting functionality through functions in the
IGraphM`Progress`
context.
Changed:
IGConnectedQ
andIGWeaklyConnectedQ
now consider the null graph to be disconnected; this is consistent with other functions such asIGTreeQ
.IGAveragePathLength
now has a"ByComponents"
option, controlling the handling of disconnected graphs.- Centrality functions:
IGCloseness
now computes the normalized closeness, i.e. the inverse of the mean distance to other vertices, by default. UseNormalized -> False
to get the previous behaviour.IGCloseness
now uses the distances only to reachable vertices when computing the closeness. In undirected disconnected graphs, it effectively computes the closeness per component. For isolated vertices (or sinks in directed graphs) it now returnsIndeterminate
.IGBetweenness
andIGBetweennessCentralization
no longer uses theMethod
option. The calculations are always fast and precise. The precision has been improved.IGBetweennessEstimate
,IGEdgeBetweennessEstimate
andIGClosenessEstimate
have been renamed toIGBetweennessCutoff
,IGEdgeBetweennessCutoff
andIGClosenessCutoff
.
IGRelativeNeighborhoodGraph
now assumes theβ -> 2
,β < 2
limit instead ofβ = 2
.IGGirth
now returnsInfinity
for the null graph.IGDiameter
now returnsIndeterminate
for the null graph.IGChordalQ
,IGChordalCompletion
andIGMaximumCardinalitySearch
now support non-simple graphs.IGReingoldTilford
andIGReingoldTilfordCircular
use a new automatic root selection algorithm. The root selection heuristic may change in the future without notice. Specify roots manually for a consistent result.IGPotentiallyConnectedQ
no longer supports directed sequences. This feature was flawed in 0.5. It may be re-added in a future version.IGLayoutKamadaKawai
andIGLayoutKamadaKawai3D
perform more iterations by default, and produce more pleasing layouts.IGPersonalizedPageRank
allows specifying the reset weights as an association from vertex names to values.- LAD isomorphism functions now support self-loops.
- Motif finder functions now support size 5 and 6 undirected motifs.
- The behaviour of the random number generator is now consistent between platforms, meaning that with a given seed, a randomized IGraph/M function will return the same result on all platforms. However, result will now be different from version 0.5 for the same seed.
Fixed:
IGPageRank
andIGPersonalizedPageRank
will now warn if the calculation did not converge with the"Arnoldi"
method. This happens only in rare cases.IGPersonalizedPageRank
: the default"PRPACK"
method returned an incorrect result when the graph was not connected and the personalization vector was not uniform.- Several community detection functions did not handle zero or one-vertex graphs correctly.
IGVertexMap
would evaluate the mapped functions twice instead of once.IGMaximumCardinalitySearch
returned incorrect ranks for graphs whose vertex names differed from their vertex indices.IGDistanceWeighted
no longer fails on edgeless graphs.IGCallawayTraisGame
no longer rejects zeros in the preference matrix.- An error would be triggered when some functions returned a zero-size matrix.
- Fixed a memory leak in the Nauty format reader.
IGMotifsVertexParticipation
would fail with Mathematica 12.2 or later.- Fixed a conflict with Mathematica 13.0's built-in isomorphism functions which could lead to a crash on Linux.
Other:
- IGraph/M now requires Mathematica 11.0 or later; on the Raspberry Pi it requires Wolfram Engine 12.2 or later.
- More robust error handling: when certain serious errors occur in the igraph C library, the Mathematica kernel is no longer forced to shut down.
- IGraph/M got leaner: smaller binary sizes.
- Documentation improvements.
v0.5.1
v0.5.1
Changes to exsiting functions:
IGReverseGraph
now supports non-simple graphs.IGLuneBetaSkeleton
is now interruptible for β < 1.
Fixes:
- Several functions failed to check if the graph was mixed (having both directed and undirected edges). This is now corrected.
IGLayoutBipartite
now uses partitions which are consistent withIGBipartitePartitions
.
Other changes:
- Many corrections and improvements to the documentation.
v0.5
v0.5
New functions:
IGSplitQ
recognizes split graphs and their degree sequences.IGThresholdQ
recognizes threshold graphs and their degree sequences.IGPotentiallyConnectedQ
recognizes the degree sequences of connected graphs.IGBigraphicalQ
recognizes the degree sequence pairs of bipartite graphs.IGEulerianQ
tests if a graph has an Eulerian path;IGEulerianPath
andIGEulerianPathVertices
find it.
Changes to existing functions:
IGGraphicalQ
andIGRealizeDegreeSequence
now support some types of non-simple graphs.IGRealizeDegreeSequence
now takes arguments in the orderindegrees
,outdegrees
for consistency with other functions (previously it wasoutdegrees
,indegrees
).IGEigenvectorCentrality
sometimes returned incorrect values for isolated vertices in weighted graphs.IGColoredSimpleGraph
no longer discards vertex names.IGModularity
now supports directed graphs.IGModularity
andIGCommunitiesMultilevel
now have a resolution parameter.IGAdjacencyMatrixPlot
now allowsNone
to be specified as the colour representing non-existing edges.IGEigenvectorCentrality
assumes the adjacency matrix of undirected graphs to have twice the number of self-loops for each vertex on the diagonal. This makes the results consistent between an undirected graph and its directed equivalent when each edge is replaced by a mutual edge pair.
Fixes:
IGLayoutReingoldTilford
no longer flips the layout.IGLayoutReingoldTilford
no longer draws overlapping tree branches.IGBarabasiAlbertGame
now allows negative values for β.IGBetweennessEstimate
sometimes returned incorrect results with finite cutoffs. This is now corrected.IGCommunitiesLeiden
: fix incorrect results when self-loops are present.IGEigenvectorCentrality
: fix incorrect results for isolated vertices and for vertices with self-loops.IGGraphicalQ
would return incorrect results when the second argument was{}
. This is now corrected.
Other changes:
IGDegreeSequenceGame
's"ConfigurationModelSimple"
method is now much faster.- More robust error handling.
Notes:
- IGraph/M 0.5 has been tested with Mathematica 10.3 and later only. Some effort has been made to allow it to work with Mathematica 10.0.2, but it has not been tested and compatibility is not guaranteed.
v0.4
v0.4
New functions:
- Deterministic graph generators:
IGKautzGraph
,IGCompleteGraph
,IGCompleteAcyclicGraph
,IGDeBruijnGraph
,IGChordalRing
,IGEmptyGraph
,IGRealizeDegreeSequence
,IGFromPrufer
,IGToPrufer
,IGKaryTree
,IGSymmetricTree
,IGBetheLattice
,IGTriangularLattice
,IGMycielskian
,IGExpressionTree
,IGShorthand
,IGFromNauty
. - Random graph generators:
IGWattsStrogatzGame
,IGCallawayTraitsGame
,IGEstablishmentGame
,IGTreeGame
,IGErdosRenyiGameGNM
,IGErdosRenyiGameGNP
. - Weighted graph functions:
IGWeightedSimpleGraph
,IGWeightedUndirectedGraph
,IGWeightedVertexDelete
,IGWeightedSubgraph
,IGUnweighted
,IGDistanceWeighted
,IGWeightedAdjacencyGraph
,IGVertexWeightedQ
,IGEdgeWeightedQ
,IGVertexStrength
,IGVertexInStrength
,IGVertexOutStrength
. - Community detection:
IGCommunitesFluid
,IGCommunitiesLeiden
. - Graph colouring functions:
IGVertexColoring
,IGEdgeColoring
,IGKVertexColoring
,IGKEdgeColoring
,IGMinimumVertexColoring
,IGMinimumEdgeColoring
,IGChromaticNumber
,IGChromaticIndex
,IGVertexColoringQ
. - Clique cover:
IGCliqueCover
,IGCliqueCoverNumber
. - Mesh/graph conversion:
IGMeshGraph
,IGMeshCellAdjacencyMatrix
,IGMeshCellAdjacencyGraph
. - Lattice generation:
IGLatticeMesh
,IGTriangularLattice
. - Proximity graphs:
IGDelaunayGraph
,IGGabrielGraph
,IGRelativeNeighborhoodGraph
,IGLuneBetaSkeleton
,IGCircleBetaSkeleton
. - Centralization:
IGDegreeCentralization
,IGBetweennessCentralization
,IGClosenessCentralization
,IGEigenvectorCentralization
. - Planar graphs:
IGPlanarQ
,IGMaximalPlanarQ
,IGOuterplanarQ
,IGKuratowskiEdges
,IGFaces
,IGDualGraph
,IGEmbeddingQ
,IGPlanarEmbedding
,IGOuterplanarEmbedding
,IGCoordinatesToEmbedding
,IGEmbeddingToCoordinates
,IGLayoutPlanar
,IGLayoutTutte
. - Adjacency lists and embeddings:
IGAdjacencyList
,IGAdjacencyGraph
. - Spanning trees and other tree-related functionality:
IGSpanningTree
,IGRandomSpanningTree
,IGSpanningTreeCount
,IGUnfoldTree
,IGTreeQ
,IGForestQ
,IGTreelikeComponents
,IGTreeGame
,IGStrahlerNumber
,IGOrientTree
. - Matching functions:
IGMaximumMatching
,IGMatchingNumber
. - Dominance:
IGDominatorTree
,IGImmediateDominators
- Maximum flow:
IGMaximumFlowValue
,IGMaximumFlowMatrix
. - Isomorphism:
IGGetIsomorphism
andIGGetSubisomorphism
(they work with multigraphs),IGColoredSimpleGraph
(for transforming multigraph isomorphism to coloured graph isomorphism). - Transitivity:
IGVertexTransitiveQ
,IGEdgeTransitiveQ
,IGSymmetricQ
,IGDistanceTransitiveQ
. - Regular graphs:
IGRegularQ
,IGStronglyRegularQ
,IGStronglyRegularParameters
,IGDistanceRegularQ
,IGIntersectonArray
. - A framework for easy property transformations and graph styling:
IGVertexProp
,IGEdgeProp
,IGEdgeVertexProp
,IGVertexMap
,IGEdgeMap
,IGVertexPropertyList
,IGEdgePropertyList
. - Import functions:
IGImport
,IGImportString
,$IGImportFormats
; support for importing Graph6, Digraph6 and Sparse6. - Export functions:
IGExport
,IGExportString
,$IGExportFormats
; support for exporting standards-compliant GraphML that can be read by other igraph interfaces (R, Python). - Matrix functions:
IGZeroDiagonal
,IGTakeUpper
,IGTakeLower
,IGAdjacencyMatrixPlot
,IGKirchhoffMatrix
,IGJointDegreeMatrix
. - Bipartite graphs:
IGBipartiteIncidenceMatrix
,IGBipartiteIncidenceGraph
,IGBipartiteProjections
. - Random walks:
IGRandomEdgeWalk
,IGRandomEdgeIndexWalk
- Connectivity:
IGGiantComponent
,IGMinimalSeparators
,IGFindMinimumCuts
,IGFindMinimalCuts
,IGBridges
,IGConnectedComponentSizes
,IGWeaklyConnectedComponentSizes
,IGGomoryHuTree
. - Efficiency measures:
IGGlobalEfficiency
,IGLocalEfficiency
,IGAverageLocalEfficiency
. - Neighbour degrees:
IGAverageNeighborDegree
,IGAverageDegreeConnectivity
- Other testing functions:
IGTriangleFreeQ
,IGSelfComplementaryQ
,IGCactusQ
,IGNullGraphQ
,IGHomeomorphicQ
,IGAdjacentVerticesQ
. - Other new functions:
IGNeighborhoodSize
,IGCoreness
,IGVoronoiCells
,IGSmoothen
,IGSimpleGraph
,IGPartitionsToMembership
,IGMembershipToPartitions
,IGSinkVertexList
,IGSourceVertexList
,IGIsolatedVertexList
,IGReorderVertices
,IGTakeSubgraph
,IGDisjointUnion
,IGTryUntil
,IGVertexAssociate
. - Added
IGIndexEdgeList
for retrieving the edge list of a graph in terms of vertex indices. This function is very fast and returns a packed array. It facilitates the efficient implementation of graph processing functions in pure Mathematica code, or interfacing with C libraries.
Updates to existing functions:
- Community detection: Several functions support the
"ClusterCount"
option now. IGRewireEdges
now supports rewiring only the start or endpoint of directed edges (instead of both).IGBipartiteQ
now supports checking that a given partitioning is valid for a bipartite graph.IGBipartitePartitions
now provides control over the ordering of partitions using its second argument.- Isomorphism functions now ignore the directedness of empty graphs.
- Isomorphism functions can now take vertex or edge colours from graph attributes.
IGBlissCanonicalGraph
will now include include vertex colours into its output when appropriate, encoded as the"Color"
vertex property.IGDistanceCounts
now optionally takes a list of starting vertices.IGBetweenness(Estimate)
andIGCloseness(Estimate)
now optionally take a set of vertices to do the calculation on.IGBetweenness(Estimate)
andIGEdgeBetweenness(Estimate)
now have theNormalized
option.IGRandomWalk
now supports edge weights and theEdgeWeights
option. UseEdgeWeights -> None
to ignore weights, and thus restore the previous behaviour.- Clustering coefficient functions now support the
"ExcludeIsolates"
option.
Incompatible changes from IGraph/M 0.3:
- A flat namespace structure is used. Functions from
IGraphM`Utilities`
have been moved toIGraphM`
. - Renamed
"MultipleEdges"
option toMultiEdges
for convenient typing and auto-completion. - Renamed
IGMinSeparators
toIGMinimumSeparators
. - Renamed
IGMakeLattice
toIGSquareLattice
. The nameIGMakeLattice
works, but it is deprecated.
Other changes:
- Improved compatibility with Mathematica versions 11.2–12.1; handling of
TwoWayRule
as an edge specification and support for edge tagged graphs. - Bug fixes, performance improvements, documentation updates, and general polish.
Notes:
- IGraph/M 0.4 will be the last version of IGraph/M to support Mathematica 10.0. Starting with IGraph/M 0.5, it will only work with versions of Mathematica that are still supported by Wolfram Research.