In my single process I have a lognormal processing time and a failure with negexp distribution. I am replicating a model that was made years ago with Arena, however I do not have the original models and the missing information can't be found. The only missing information is how the randomness with Arena was used and though I'd like to variate the random seed to obtain a close output (it should be around 13.7k, whereas now I get 15k). I do not know how the RandomSeed affects mathematically these two distributions and it would be really useful to know it. Does anyone knows it?
Thanks in advance for the answers
Solved! Go to Solution.
with the two distributions you are referring to your "lognormal processing time and your failure with negexp distribution"? Also, as you are talking about stochastic models and their comparison, you should not consider an exact comparison of your simulation results but rather look at their standart deviation and variance. If that is roughly the same in both models, like it might be with the values that you posted, then it means that both your models give similar results. This is what you are normally getting with two different stochastic simulations modelling correctly the same situation.
Yes I would suggest doing so as well, you can double click the Eventcontroller and under Tools you will find the Random Numbers Variant. Here you can try out different values an see when your distribution gets closer to the desired value.