Reproducible simulation?

[bryo-han]

@gregorgorjanc you have set up the simulation code for a small dataset and that includes saving the R session's image. This is all good.

I was wondering if these simulation could be made reproducible. Just using set.seed did not help, the resulting output files differed between individual simulation runs. Do you happen to know why?

Feel free to close as it's not our priority to make this work. And this behaviour of ASR may be documented elsewhere?

[gregorgorjanc]

@hannesbecher this is a known issue and I think it's coming from runMacs() (sending seed to MaCS does not have the effect we want) - can you check if we are using it? We could swap runMacs() with quickHaplo() in this simple simulation.

[hannesbecher]

Makes sense, looking into it!