graphstream/gs-core

Reproducibility of random processes

Closed this issue · 1 comments

Hi,
is there anyway to seed a generator in such a way that, given a seed, it always generates the same graph?
This is extremely important in order to use graphstream for scientific simulations, as the same runs need to be reproducible.

I believe BaseGenerator should be modified with a two new constructors accepting respectively:

  • a RandomGenerator (possibly from Apache Commons Math, but a Java Random could work as well); or
  • an int representing the seed for the internal Random.

Nevermind, I missed setRandomSeed within BaseGenerator. Please consider this issue invalid.