A C++ implementation of the LSD algorithm to detect super bubbles.
To install:
- Clone git repro.
- Run configure.
- Run make.
git clone git@github.com:Fabianexe/clsd.git
cd clsd
./configure
make
clsd needs an edge list as input file. The format is that in every line is one edge, where the labels of the vertices are seperated by a space or tab. As an example, a graph with the directed edges: 1->2, 1->3, 2->4, and 3->4 would have this file:
1 2
1 3
2 4
3 4
The file could contain multi-edges but the parser ignores them because they have no information of superbubbles. Many edgelist writes put properties behind this for example an edge 1->2 with a weight 0.2 would look like:
1 2 0.2
This is also possible but again clsd ignores the weights.
clsd can compute different stastistics. Note that clsd ignores multi-edges. Thus every stastistic is done by ignoring multiple appearence of the same edge. The output is a list of numbers sperated by &. Before content line is the header line. The header shorts mean:
| Metric | Short | Paper | Description |
|---|---|---|---|
| Vertices | N | - | Number of Vertices |
| Edges | M | - | Number of Edges |
| Multi Edges | ME | - | Fraction of Edges that are multi edges |
| Max Degree | deg | - | The largest degree |
| Graph Density | GD | - | (2m)/(n(n-1)) |
| Connected components | CC | - | The number of connected components |
| Assortativity | R | Mixing patterns in networks | Valiation of euqal degrees on neighbor vertices |
| Self-Similarity | SS | Towards a Theory of Scale-Free Graphs: Definition, Properties, and Implications | This is maximized when high-degree nodes are connected to other high-degree nodes |
| Metric | Short | Paper | Description |
|---|---|---|---|
| Max in Degree | - | The largest inDegree | |
| Max out Degree | - | The largest outDegree | |
| Strongly connected components | SCC | - | The number of strongly connected components |
| Bidirectional edges | BE | - | Fraction of the edges that have invertet edge in Each |
| Directed assortativity | Edge direction and the structure of networks | Valiation of euqal degrees on neighbor vertices by comparing in- to in-degree | |
| Directed assortativity | Edge direction and the structure of networks | Valiation of euqal degrees on neighbor vertices by comparing in- to out-degree | |
| Directed assortativity | Edge direction and the structure of networks | Valiation of euqal degrees on neighbor vertices by comparing out- to in-degree | |
| Directed assortativity | Edge direction and the structure of networks | Valiation of euqal degrees on neighbor vertices by comparing out- to out-degree | |
| Heterogeneity index | H | Entropy and Heterogeneity Measures for Directed Graphs | Entropy of neighbor degrees |
| Metric | Short | Description |
|---|---|---|
| Number of Superbubbles | S | How many superbubbles in the graph |
| Vertices Superbubbles | VS | The fraction of vertices that are part of a superbubble |
| Edges Superbubbles | ES | The fraction of edges that are part of a superbubble |
| Number of Mini-Superbubbles | MS | How many mini superbubbles in the graph i.e. superbubbles with two vertices |
| Vertices Complex Superbubbles | VCS | The fraction of vertices that are part of a complex superbubble |
| Edges Complex Superbubbles | ECS | The fraction of edges that are part of a complex superbubble |
| Max Vertices | mVS | What is the maximum of vertices in one superbubble |
| Max Edges | mES | What is the maximum of edges in one superbubble |
| Superbubble Complex | C | How many non-overlapping superbubble complexes exist |
| Superbubble Complex size | CS | The maximum number of superbubbles in one complex |
| Deepest Superbubble | depth | The maximal depth that a superbubble have |
| maximal paths in Superbubble | P | The maximal number of a paths in one superbubble |
| longest path in Superbubble | PL | The maximal length of a path in a superbubble |
| Metric | Short | Description |
|---|---|---|
| Average number of paths | aP | The average number of paths in non-mini superbubbles |
| Average path length | aPL | The average path length in non-mini superbubbles |
| Superbubble Density | SD | The average value of (2m)/(n(n-1)) for all superbubbles in non-mini superbubbles |
