About the Code
The code implements the Anchored Vertex Tracking in Dynamic Social Networks based on the order-based core maintenance algorithm.
File Structure
The code consists of
- mainavt.cc: the main program
- defs.h: this file defines some useful macros
- core.h: this header file defines the base class for core maintenance algorithm
- gadget: this folder contains gadgets for reading files and implements several data structures
- disjoint.h, disjoint.cc: a union-find-delete data structure with constant time deletion complexity
- gadget.h, gadget.cc: for reading data files, the format of which will be explained in details later
- heap.h, heap.cc: a minimum heap
- sb_tree.h, sb_tree.cc: the size-balanced tree
- treap.h, treap.cc: the treap (tree + heap)
- glist: this folder contains the code for the order-based core maintenance algorithm
- incavt.cpp: including the main functions of AVT algorithm
- glist.cpp: including the used data stucture for k-core maintenance and functions for k-core computation
Command
There are five options to run the code.
- -p: the path of the initial graph file
- -A: the allocated size
- -T: the number of dynamic graphs (or graph snapshots)
the edge insert and edge remove file is indexed by the index of snapshots.
All the files with edge insert and edge remove are stored in the directory ./dataset/edge_insert/ and ./dataset/edge_remove/.
The file name is formated as graphname_1,graphname_2, ....
- -F: the graph file name
- -K: the core number
- -R: the edge remove directory,remember that the graph file name and the file name in this directory should follow the format: graph_name_t.txt
- -I: the edge insert directory,remember that the graph file name and the file name in this directory should follow the format: graph_name_t.txt
An example command: ./incavt -p dataset/(dataset_name)/(dataset_name)_graph.txt -R dataset/(dataset_name)/(dataset_name)_edge_remove/ -I (dataset_name)/Email/(dataset_name)_edge_insert/ -A 2 -F (dataset_name)_graph -T 4 -K 5
Compiling the Code
Use Makefile included and notice that in default, -O2 optimization option is triggered and the code is compiled based on C++11.