replication - Netlogo - How to replicate an experiment adjusting one parameter using import-world? -

i have simulation in agents move around physical grid exchange ideas on topics neighbours. have conducted study , replicate 1 modification - adjusted "transfer" parameter ideas exchange @ slower rate.

to try , perfect replication have used export/import world procedure. when re-run simulation tick 0 imported world , none of parameters changed, model identical. same output, agents move , have same interactions , identical. great

however, when re-run simulation tick 0 same imported world, adjust 1 paramater (transfer_rate), agents move around physical space differently, meet different agents original simulation , different results.

perhaps adjusting 1 parameter pseudorandom number has altered somehow? know of way of having controlled , constant exception of 1 transfer_rate parameter adjustment? perhaps there simple solution.

thank time everyone.

two possibilities might help. try with-local-randomness isolate random processes affected change in transfer_rate. alternatively, 1 simulation doesn't tell anyway, move away trying match dynamics , run multiple simulations , compare 2 sets of simulations.