Flatten nested DMN tables

Hey,

In Camunda Modeler I have nested DMN Tables. and I want to flatten them to have a unique table covering all values of input and outputs. I’m searching for a feature “Reflect Decision” but i don’t see it on my modeler. Could you please help me to figure out how to achieve this need?

Thanks in advance

Hi @Saloua_Bennani,

Please tell me more about your motivation. Why do you want to achieve this?
To combine two decision tables t1 and t2, where t2 requires the output of t1. You would need to add all the inputs of t1 to t2 and replace a value of t1 in a rule of t2, with all rules in t1 that produce this value. This may require duplicating the rule and this assumes that both decision tables are unique.

I know about one research paper and prototype that may help. The authors transform any DMN decision into a unique one:
https://www.researchgate.net/publication/325784947_Disambiguation_of_DMN_Decision_Tables

1 Like

Hi, Thank you for your repsonse.
For me it’s not about disambiguating rules. It’s more that i have sub decision tables and i need to have an overall of all potential combinations. I want to end up with a file containing all possible combinations of my input columns ans their associated outputs.

For example :

my sub DMN table 1 has:
input 1|output1

my sub DMN table 2 has
input2 |output2 which equals input 1.

So whenever I have a value for input & or input 2, i can know the output1.
I hope my example is clear and you can help me conducting it on the modeler.

Hi @Saloua_Bennani,

I understand that you consider input1 and input2 to be alternatives in the final decision table. Is that correct? In this case, you do not even need to replicate the tables, you can simply append on to the other (and change the hit policy to first):

Is this what you want to achieve?

Thank you @StephanHaarmann.

Yeah this is it. but how did you do to create and visualize the rules of the table “combined decision”. I can not do it manually since on the real example I have many rows and combinations. It would be error prone.

Thank you for your input on that.