question

yoni A avatar image
0 Likes"
yoni A asked yoni A commented

how to load an Empirical distribution?

after collecting sufficient cycle time data of a process, I use experfit to evaluate and determine the best distribution to fit the data, the results says it is indeterminate, does that mean that there is no distribution that explains the data collected?

the software suggest you use an empirical distribution, to load into the processor's process time, how do I load the results of the empirical distribution into the processor?

FlexSim 18.1.0
distributions
· 8
5 |100000

Up to 12 attachments (including images) can be used with a maximum of 23.8 MiB each and 47.7 MiB total.

Jeff Nordgren avatar image Jeff Nordgren commented ·
@yoni A

You will probably need to send in your model or a sample model of the problems so that we can take a look to see what is happening within the model. If the data returns an indeterminate response from OptQuest then maybe the parameters were not entered in the right way? Because it seems to me there should always be some kind of distribution for any set of input, unless the input in incomplete or not entered correctly.

0 Likes 0 ·
yoni A avatar image yoni A Jeff Nordgren commented ·

Jeff see attached file @Jeff Nordgren, under the title of the question

0 Likes 0 ·
yoni A avatar image yoni A commented ·

@Jeff Nordgren

see attached file, I couldn't attache dthe expert fit file since it is not accepted

statistical-distributions.xlsx

0 Likes 0 ·
Ben Wilson avatar image Ben Wilson ♦♦ yoni A commented ·

@yoni A,

Sorry you couldn't attach the ExpertFit file.

I have enabled .efp file uploads. If your file extension was something else, let me know.

0 Likes 0 ·
yoni A avatar image yoni A Ben Wilson ♦♦ commented ·

project-2-real-numbers.efp

@Ben Wilson

see experfit file

0 Likes 0 ·
Show more comments
yoni A avatar image yoni A commented ·
0 Likes 0 ·

1 Answer

Jordan Johnson avatar image
1 Like"
Jordan Johnson answered Jeff Nordgren edited

@yoni A, I am not too familiar with ExpertFit specifically, so I don't know when it generates that error. However, I am familiar with curve fitting in general. Basically, you find the best fit from a given distribution, and then give it a "goodness of fit" score. My guess is that if that score is not good enough for any distribution, ExpertFit will not recommend a distribution to fit the data. If you make a histogram of the data, are there multiple modes? Most distributions that ExpertFit attempts to fit can only handle a single mode.

5 |100000

Up to 12 attachments (including images) can be used with a maximum of 23.8 MiB each and 47.7 MiB total.