is it possible to have a mean-minimum-maximum distribution?

[oliver_h]

[ FlexSim 17.0.2 ]

Hi,

I have some data for a process, who has different process times each process.

so the mean process time is 5 hours, but the process has at least 2 hours and 7 hours at the most. so the distribution is not symmetrical.

a normal distribution is symmetrical, so i can't use it. a uniform distribution is from 2 to 7, but the mean value can be anthing in between.

does anyone have a solution for this? thank you in advance!

[Matthew_Gillespie · Accepted solution]

Yes, you can use the triangular distribution. You just need to specify the min, max, and mode.

[oliver_h]

@►Matthew Gillespie

Thank you for your fast answer. so "mode" in that case has the same meaning as "mean", am i right?

thanks anyways, you helped a lot.

[Matthew_Gillespie]

You can also use ExpertFit, a curve fitting program included with FlexSim, to fit a distribution to your data. You can find it in the Statistics menu.

[Matthew_Gillespie]

(Sorry if this is isn't what you were asking) Mode and mean are both "averages", but they are a little bit different.

Mean is the traditional average everyone thinks of :

mean = a1 + a2 + ... an / n

Mode is the most repeated value:

1, 2, 3, 1, 2, 1	//mode = 1, mean = 1.67

That said, the triangular distribution is already fairly inaccurate so it probably doesn't matter too much which one you use. The mode value really just needs to represent the most likely value, so however you want to represent that.

[cliff_king]

Because (min + max + mode)/3 = avg, you can back calculate the mode as mode = (3 * avg) - min - max