Value in an Empirical distribution

[allen_greenwood]

[ FlexSim 21.2.2 ]

How to get the value of a particular row and column in an Empirical distribution table? Seems like it should be able to be referenced like a global table but that doesn't seem to work.

[ryan_c10]

Hello @Allen Greenwood,

That's not quite how Empirical Distributions work. They work like using any other statistical distribution: you define the parameters of the distribution and FlexSim outputs a random value that fits the distribution every time the distribution is accessed. The Empirical Distribution works the same way, except you manually define the parameters.

Does that answer your question?

[moehlmann_fe · Accepted solution]

The data of an empirical distribution is saved in a treenode containing bundle data in the tree of the respective "empirical distribution" object.

Values in it can be accessed with the "getbundlevalue" command.

// Get reference to the node containing the data - 
// replace "EmpiricalDistribution1" with the name of the distribution
treenode bundle = Model.find("/Tools/EmpiricalDistributions/EmpiricalDistribution1>variables/data");

// Assign value to variable
double val = getbundlevalue(bundle, <row>, <column>);

Important note: The rows ("entry") and columns ("field") of bundles are 0-indexed.

[Jeanette_Fullmer]

Hi @Allen Greenwood, was one of Ryan Clark's or Felix Möhlmann's answers helpful? If so, please click the "Accept" button at the bottom of the one that best answers your question. Or if you still have questions, add a comment and we'll continue the conversation.

If we haven't heard back from you within 3 business days we'll auto-accept an answer, but you can always unaccept and comment back to reopen your question.

[allen_greenwood]

Thanks Felix. I didn't realize it was saved as a bundle. In code, I need to know how many rows are in the data. In my case, I am specifying an empirical distribution using pairs of values and corresponding percents. Whereas the new Empirical Distribution is more for a set of raw data. It seems that the older approach of using a global table and dempirical (or cempirical) distribution is better for what I need.

[JordanLJohnson]

If you are using pairs of values and corresponding percents, use the "weighted" option and put your percents as the weights. Then I think you'll have what you are looking for.

[allen_greenwood]

I did that initially. However, in some code in the model, I need to know how many rows are in the distribution table. That's were I got stuck since I couldn't just use a table method.

[michael_d11]

Hi, is there a tutorial showing how to use this approach (with weights)?