Evaluate link parition

This program evaluate a link partition of a graph according to different quality functions. This program process only undirected unweighted edges. The quality functions that can be computed are:

Partition density. "Link communities reveal multiscale complexity in networks" Y.-Y. Ahn, J.P. Bagrow, S. Lehmann
Linegraphe Modularity. "Line Graphs, Link Partitions and Overlapping Communities" T.S.Evans, R.Lambiotte
Expected Nodes. "Expected Nodes: a quality function for the detection of link communities" N. Gaumont, F. Queyroi, C. Magnien, M. Latapy

Author

Author : Gaumont Noé
Email : noe.gaumont@lip6.fr
Location : Paris, France

####Version 1.0 Disclaimer

If you find a bug, please send a bug report to noe.gaumont@lip6.fr including if necessary the input file and the parameters that caused the bug. You can also send me any comment or suggestion about the program.

Compilation

This program needs boost to be installed. To install boost:

On debian : sudo apt-get install libboost-all-dev
On archlinux : sudo yaourt -S boost
See http://www.boost.org

Once boost installed, just type:

make evaluation

If you don't want to bother with boost. Use the care achive "builder.bin":

./builder.bin # this creates builder/
./builder/re-execute.sh # this compile everything
mv builder/rootfs/home/gaumont/CodingPlace/linkStream/evaluation .

After the compilation, you are ready to use the program. See the Usage section.

Usage

Here are typical usages :

./evaluation -i INPUTFILE -q 1,2,3,4

This evaluates the graph stored in INPUTFILE with the quality functions 1,2,3 and 4.

If you don't like ID, you can also use names:

./evaluation -i INPUTFILE -q ExpectedNodes,PartionDensity,E1,E2,E3

Of course, you can use only one quality function:

./evaluation -i INPUTFILE -q ExpectedNodes

INPUTFILE

The input file should contain a list of undirected unweighted links with a community. A links with a community should be in the format: Source, Target, Community. For a triangle with 2 communities:

1 2 1
1 3 1
2 3 2

The entry could be space or tab separated. You can generate toy example entry file. See Examples generation.

Arguments

With the option q, you can provide which quality function has to be tested. The quality functions has to be provided as a comma separated list of their id numbers or names. Available quality function are:

Parition Density introduced by Ahn et al. ID=2, name=ParitionDensity
The first version of the quality function introduce by Evans et al. . id=3 , name=E1
The second version of the quality function introduce by Evans et al. . id=4 , name=E2
The third version of the quality function introduce by Evans et al. . id=5 , name=E3
Expected Nodes Id=1 , name= ExpectedNodes

Please refer to the according paper to find their definition.

Examples generation

In the folder Example, there is a script to generate toy examples. The script creates a complete graph of given size with a link partition. To create the link partition, the script need 2 parameters. The first one is a number of nodes and the second is the number of groups (1,2 or 3). Please refere to the paper "ExpectedNodes :[..]" to have more information.

./1cliqueGeneration.sh 100 15 3 > example1.txt

License

MIT

ngaumont / ExpectedNodes