Truncated(bounded) distribution in Arena.

[Hirotaro]

Have any of you used a truncated distribution in your Arena model? While some other simulation software enables an user to easily implement truncated distributions say:

Expo(1) truncated(bounded) at 100
NORM(0, 5) bounded at <0,100>,
it is apparently not the case with Arena.
What is the best (most efficient) way to implement truncated distributions in Arena?
Looking forward for your comments.

[ewilliams]

Perhaps the easiest alternative is to use the Expression table in the Advanced Process template. Open the builder and give your expression a name you like, like ExamTime. Leave rows and columns blank for a simple scalar expression. For Expression Value, type, for example, MAX(EXPO(1),100). Then use ExamTime in your model wherever you like.

[Hirotaro]

Thank you for your comment E.Williams.

I quess that other option can be to simply resample.

If the bounded value becomes the sample value (as in your option) and a big share of CDF lies outwith the upper and lower bound then the resulting distribution can be incorrect (depending on the model settings). For example, if one would use normal distribution with mean=0, and lower bound at 0, then 50% of values would be equal to 0.

Moreover 2 bounds require another trick while using your option without resampling.