calculateDescriptors: Generalized descriptor calculation

In QuACN: QuACN: Quantitative Analysis of Complex Networks

Description

The method calculates multiple descriptors for a list of graphs.

Usage

calculateDescriptors(graphs, ..., labels=FALSE, log=FALSE)

Arguments

graphs	either a list of or a single graphNEL object.
...	descriptors to calculate and arguments to pass, see ‘Details’.
labels	whether or not the columns of the resulting data frame should be named using the getLabels() method.
log	whether or not informative messages about the progress of the calculation should be printed

Details

calculateDescriptors() calls each function specified in ‘...’ for every graph in the given list and creates a data frame containing the calculated data. You can specify the functions either as strings (such as “totalAdjacency”) or using the numbers from the following table (e.g., 2001). For convenience, the multiples of 1000 denote entire groups of descriptors.

1000	--- all of 1xxx
1001	wiener
1002	harary
1003	balabanJ
1004	meanDistanceDeviation
1005	compactness
1006	productOfRowSums
1007	hyperDistancePathIndex
1008	dobrynin
2000	--- all of 2xxx
2001	totalAdjacency
2002	zagreb1
2003	zagreb2
2004	modifiedZagreb
2005	augmentedZagreb
2006	variableZagreb
2007	randic
2008	complexityIndexB
2009	normalizedEdgeComplexity
2010	atomBondConnectivity
2011	geometricArithmetic1
2012	geometricArithmetic2
2013	geometricArithmetic3
2014	narumiKatayama
3000	--- all of 3xxx
3001	topologicalInfoContent
3002	bonchev1
3003	bonchev2
3004	bertz
3005	radialCentric
3006	vertexDegree
3007	balabanlike1
3008	balabanlike2
3009	graphVertexComplexity
3010	informationBondIndex
3011	edgeEqualityMIC
3012	edgeMagnitudeMIC
3013	symmetryIndex
3014	bonchev3
3015	graphDistanceComplexity
3016	distanceDegreeMIC
3017	distanceDegreeEquality
3018	distanceDegreeCompactness
3019	informationLayerIndex
4000	--- all of 4xxx
4001	mediumArticulation
4002	efficiency
4003	graphIndexComplexity
4004	offdiagonal
4005	spanningTreeSensitivity
4006	distanceDegreeCentric
4007	distanceCodeCentric
5000	--- all of 5xxx
5001	infoTheoreticGCM: vertcent, exp
5002	infoTheoreticGCM: vertcent, lin
5003	infoTheoreticGCM: sphere, exp
5004	infoTheoreticGCM: sphere, lin
5005	infoTheoreticGCM: pathlength, exp
5006	infoTheoreticGCM: pathlength, lin
5007	infoTheoreticGCM: degree, exp
5008	infoTheoreticGCM: degree, lin
5009	infoTheoreticLabeledV1: exp
5010	infoTheoreticLabeledV1: lin
5011	infoTheoreticLabeledV2
5012	infoTheoreticLabeledE: exp
5013	infoTheoreticLabeledE: lin
6000	--- all of 6xxx
6001	eigenvalueBased: adjacencyMatrix, s=1
6002	eigenvalueBased: adjacencyMatrix, s=2
6003	eigenvalueBased: laplaceMatrix, s=1
6004	eigenvalueBased: laplaceMatrix, s=2
6005	eigenvalueBased: distanceMatrix, s=1
6006	eigenvalueBased: distanceMatrix, s=2
6007	eigenvalueBased: distancePathMatrix, s=1
6008	eigenvalueBased: distancePathMatrix, s=2
6009	eigenvalueBased: augmentedMatrix, s=1
6010	eigenvalueBased: augmentedMatrix, s=2
6011	eigenvalueBased: extendedAdjacencyMatrix, s=1
6012	eigenvalueBased: extendedAdjacencyMatrix, s=2
6013	eigenvalueBased: vertConnectMatrix, s=1
6014	eigenvalueBased: vertConnectMatrix, s=2
6015	eigenvalueBased: randomWalkMatrix, s=1
6016	eigenvalueBased: randomWalkMatrix, s=2
6017	eigenvalueBased: weightStrucFuncMatrix_lin, s=1
6018	eigenvalueBased: weightStrucFuncMatrix_lin, s=2
6019	eigenvalueBased: weightStrucFuncMatrix_exp, s=1
6020	eigenvalueBased: weightStrucFuncMatrix_exp, s=2
6021	energy
6022	laplacianEnergy
6023	estrada
6024	laplacianEstrada
6025	spectralRadius
7000	--- all of 7xxx
7001	oneEdgeDeletedSubgraphComplexity
7002	twoEdgesDeletedSubgraphComplexity
7003	globalClusteringCoeff
8000	--- all of 8xxx
8001	connectivityID
8002	minConnectivityID
8003	primeID
8004	bondOrderID
8005	balabanID
8006	minBalabanID
8007	weightedID
8008	huXuID

The arguments to these functions, such as the distance matrix or the list of vertex degrees, will be automatically supplied and reused. After each function (or group of functions), regardless of whether it was referred to by name or by its assigned number, you may optionally pass extra arguments as a list, but note that this will not override the calculated arguments. If you wish to pass the same extra arguments to multiple functions, you can concatenate the latter to a vector.

When functions are given by name, an “@NAME” suffix can be used to give the column a different name in the output data frame. This is needed when you want to calculate a descriptor more than once with varying arguments.

If log is TRUE, a progress message is printed to the standard output connection for each graph in the list.

Value

A data frame where rows and columns represent the input graphs and the desired descriptors, respectively. The rows will be named according to the graph list; the column names are the names of the called functions if labels is FALSE, otherwise the label expressions as returned by getLabels() (and found in the vignette).

Author(s)

Michael Schutte

Examples

library(RBGL)
set.seed(123)
g <- randomGraph(1:8, 1:5, 0.36, weights=FALSE)

calculateDescriptors(g, 1000, 2002, 2003)

calculateDescriptors(g, "randic", "offdiagonal", 7000, labels=TRUE)

# these will give the same results (although named differently):
calculateDescriptors(g, c(6011, 6013), list(s=3))
calculateDescriptors(g,
  "eigenvalueBased@ea", list(matrix_function="extendedAdjacencyMatrix", s=3),
  "eigenvalueBased@vc", list(matrix_function="vertConnectMatrix", s=3))

QuACN documentation built on May 2, 2019, 5:46 p.m.