|
Usage:
NetworkEvolution.jar <job> <job_parameters>
Jobs:
- simulate - Simulate Network Evolution
- path - MCMC for Path Sampling
- gibbs - Gibbs Sampler for Parameter Estimation
Job Parameters:
Job: simulate
- Evolution Model - 1: Independent Edge Model, 2:
Neighbour-Dependent Model
- Directory containing network and other files (. for current
directory)
- Reference Network Filename
- Start Network Filename
- Core Hyperedges Filename ("" for none)
- Prohibited Hyperedges Filename ("" for none)
- Insertion Rate
- Deletion Rate
- Number of Simulation Runs
- Number of Iterations per Run
- Update Interval for Output Display
Job: path
- Evolution Model - 1: Independent Edge Model, 2:
Neighbour-Dependent Model
- Directory containing network and other files (. for current
directory)
- Reference Network Filename
- Start Network Filename
- End Network Filename
- Core Hyperedges Filename ("" for none)
- Prohibited Hyperedges Filename ("" for none)
- Insertion Rate
- Deletion Rate
- Number of Simulation Runs
- Number of Iterations per Run
- Number of Iterations for Burn-in Period
- Update Interval for Output Display
Job: gibbs
- 1. Evolution Model - 1: Independent Edge Model, 2:
Neighbour-Dependent Model
- Directory containing network and other files (. for current
directory)
- Reference Network Filename
- Start Network Filename
- End Network Filename
- Core Hyperedges Filename ("" for none)
- Prohibited Hyperedges Filename ("" for none)
- Insertion Rate (-ve value for random starting value)
- Deletion Rate (-ve value for random starting value)
- Number of Simulation Runs
- Number of Iterations per Run
- Number of Iterations for Burn-in Period
- Update Interval for Output Display
Examples:
Assumption: Data files are in the same directory as the .jar file
- Simulating network evolution with 10000 iterations, update every
100th iteration, 2 runs
java -jar NetworkEvolution.jar path 2 . 00030_ref.txt 00030_pae.txt core00030.txt prohibited00030.txt 0.04 0.01 2 11000 100
-
MCMC for path sampling with 11000 iterations out of which 1000 are burn-in period, update every 10th iteration, single run
java -jar NetworkEvolution.jar path 2 . 00030_ref.txt 00030_pae.txt 00030_pst.txt core00030.txt prohibited00030.txt 0.04 0.01 1 11000 1000 10
-
Gibbs Sampler with 11000 iterations out of which 1000 are burn-in period, update every 10th iteration, single run
java -jar NetworkEvolution.jar gibbs 2 . 00030_ref.txt 00030_pae.txt 00030_pst.txt core00030.txt prohibited00030.txt 0.04 0.01 1 11000 1000 10
-
Gibbs Sampler with 11000 iterations out of which 1000 are burn-in period, update every 10th iteration, random starting rates, 3 runs
java -jar NetworkEvolution.jar gibbs 2 . 00030_ref.txt 00030_pae.txt 00030_pst.txt core00030.txt prohibited00030.txt -1 -1 3 11000 1000 10
|