Using the Input Analyzer
Need to have A text file with numbers A “hard” return after each number no commas, no tabs Example 5.2 4 -100
Open Input Analyzer in Arena Folder
The application would look like this:
Go to File, New
This should happen: a file called Input1 (or Input2, or InputX)
Go to File, Data File, Use Existing
Next change the file type to txt
Next find your file
The software will automatically create a histogram
Go to Fit, select Fit All
The Input Analyzer will provide the expression for the best fit
But sometimes is better to use another distribution The most commonly used distributions for simple simulation models are Uniform Triangular Exponential Thus we can use the one (from the 3 above) with the lowest error
To view the error of all the distributions (as to select a known distribution) go to Fit All Summary
The software will show the error for all the distribution it analyzes The software will show the error for all the distribution it analyzes. In this case, out of Expo, Uniform and Triangular, the smallest error is the Triangular
So we should use the Triangular distribution So we should use the Triangular distribution. To get the parameters go to Fit, select Triangular.
Nothings shows, must click in the other Window (to make it active) Click here, in the other Window
Done Now the raw data has been characterized as a random variable. In this case a Triangular Distribution with TRIA(-0.5, 5, 12.5) Note that in this case the minimum is a negative value. Processing times or arrival times cannot be negative. Thus when implementing in Arena this needs to be modified to TRIA (0,5,12.5)