Datasets of software network structure for defect prediction.
- Came(1.0/1.4/1.6)
- Jedit(3.2/4.1/4.2/4.3)
- Lucene(2.0/2.2/2.4)
- Synapse(1.0/1.1/1.2)
- Xerces(1.2/1.3/1.4)
defect information is obtained from tera-PROMISE repository (http://openscience.us/repo/defect/)
32 software network structure features include 18 global network structure features and 14 ego network structure features. The 34 different software structure features were calculated by software source code.
1.Global network measures
- Degree Nodes adjacent to a given node
- Out degree Number of edges incoming to that node
- In degree Number of edges outgoing from that node
- Ripple degree Number of nodes that that node can reach along directed paths
- Average neighbor degree Average degree of the neighborhood of that node
- Betweenness Sum of shortest paths that pass through that node
- Pagerank Ranking of that nodes based on the structure of the incoming links
- Clustering Fraction of possible triangles through that node that exist
- Square clustering Fraction of possible squares that exist at the node
- Closeness centrality Reciprocal of the sum of the shortest path distances from that node to all other nodes
- Degree_centrality Fraction of nodes it is connected to that node
- Out degree centrality Fraction of nodes its outgoing edges are connected to that node
- In degree centrality Fraction of nodes its incoming edges are connected to that node
- Betweenness centrality Sum of the fraction of all-pairs shortest paths that pass through that node
- Load centrality Fraction of all shortest paths that pass through that node
- Number of cliques Number of maximal cliques for each node
- Core number Largest value k of a k-core containing that node
- Eccentricity Maximum distance from that node to all other nodes
2.Ego network measures
- Size Nodes that ego is directly connected to
- Ties Directed ties corresponds to the number of edges
- Pairs Unique pairs of nodes, i.e., Size×(Size−1)
- Density Ties/Pairs
- nWeakComponents Weak components in the ego network
- pWeakComponents Weak components normalized by size
- 2StepReach Nodes ego can reach within two steps normalized by Size
- 2StepPreach 2StepReach/(N-1)
- Reach_efficiency 2StepReach normalized by the sum of the size of the ego’s * * every Neighbor’s ego network
- Broker Pairs not directly connected to each other
- nBroker Broker normalized by the number of pairs
- nClose Number of closed triads ego is involved in
- EgoBetweenness Shortest paths between neighbors that pass-through ego
- nEgoBetweenness EgoBetweenness normalized by the size of the ego network