Adding decisions in Array

indranil nanda

unread,

Apr 14, 2017, 8:40:45 AM4/14/17

to OpenRules

Recently we started exploring OpenRules and was trying to make a decisionmodel based on excel only.

And we are stuck in a situation for which were not able to locate any suitable sample application in 6.4.1 download.

I want to know is it possible to put decisions from different DTs in to an array.

So what we want to achieve is, let we have possible decisions

Data DecisionImportance decisionImportances
myDecision	importance
My Decision	Importance
Decision1	9
Decision2	10
Decision3	15
Decision4	12
Decision5	6
	-99 //null or no decision

Now based on customer data(input), 1^st decision table will give Decision 1/null, 2^nd decision table will give Decision2 or null etc.

So I want to collect all these decisions in a Array and want to do some logic based on importance to provide the final decision.

//DecisionTable ViewMyShipment1
If	If	Then
var1	var2	Recommendation[0]
Yes	Yes	Decision1

//DecisionTable ViewMyShipment2
If	If	Then
Var3	Var4	Recommendation[1]
Yes	Yes	Decision2

So my question is how to write this (Then/Action/Conclusion) to put into decision array.

Believe I can do += and to concatenate all decisions in a String and then split in Java. But is there any other alternative.

OpenRules Support

unread,

Apr 14, 2017, 8:49:02 AM4/14/17

to OpenRules

You probably want to create an array of recommendations (not array of decisions). The standard workspace "openrules.dmn" provides two examples that deal with arrays of variables: DecisionAggregatedValues and DecisionUserv. It seems your example is simpler and if you may specify a more clear logic with input data and expected results of what you actually trying to achieve we will be glad to provide a solution for you.

Support

indranil nanda

unread,

Apr 14, 2017, 7:02:59 PM4/14/17

to OpenRules

Glossary glossary
Variable	Business Concept	Attribute
Var 1	Customer	var1
Var 2		var2
Var 3		var3
Var 4		var4
Var 5		var5
Var 6		var6
Var 7		var7
Var 8		var8
Var 9		var9
Var 10		var10
Recommendations	Recommendations	recommendations
Recommendation Name	RecommendationWeightage	recommendationName
Weight		weight
Recommendation Name For Consumers		recommendationNameForConsumers

Datatype Recommendations
String[]	recommendations


Data RecommendationWeightage recommendationWeightages
recommendationName	weight	recommendationNameForConsumers
Recommendation Name	Weight	Recommendation Name For Consumers
recommendation1	1	IS_ELIGIBLE_FOR_RECOMMENDATION_1
recommendation2	2	IS_ELIGIBLE_FOR_RECOMMENDATION_2
recommendation3	4	IS_ELIGIBLE_FOR_RECOMMENDATION_3



DT DeciosnForRecommendation1
If	If	Conclusion
Var1	Var2	Recommendations
A	B	recommendation1


DT DeciosnForRecommendation2
If	If	Conclusion
Var3	Var4	Recommendations
C	D	recommendation2

In Java side:

I am reading the Constant data recommendationWeightage from Excel file.

And I am expecting the recommendation[] will contain some data like [“recommendation1”, “recommendation5”, “recommendation6”] etc.

Now based on priority/weightage I am going to sort those data. And give some top 5 or top 2 record.

So I have doubt in how to add recommendations dynamically in to the array.

Cause the examples I saw, there it is part of data which is fed in to the decision model.

But what I want, is to add output of different decision tables into the array.

So believe I have to update something in the sections of DT highlighted in yellow.

-----------------------------------------------------

OpenRules Support

unread,

Apr 15, 2017, 9:35:28 AM4/15/17

to OpenRules

Hi Indranil,

We created a sample working project that (we believe) demonstrates how to resolve your problem. It may be useful for other customers as well. You may download this working project from

http://openrules.com/public/DecisionSortArray.zip

Here is the problem description. We have an array of Customers and an array of Products. Customers may like certain features and products may have (or not) these features. Each feature may have a weight/importance that contributes to the score when we assign a product to a customer. Here is an example of the scoring rules:

To execute these rules for every customer, you may use a table like this one:

The Java class Customer contains an array of all products that will be evaluated using these rules. Then we may evaluate (score) all products using Java standard method Arrays.sort(products). Note that our Java class Product implements Comparable<Product> and here is a simple compareTo() method:

public int compareTo(Product product) {

if(product.customerScore > this.customerScore)

return 1;

else

return -1;

}

If you need more complex comparison of products, this method may actually call an Excel-based decision table.

We are attaching the complete file DecisionSortArray.xls with test data: Here are the execution results for the first customer:

=== Customer: Robinson

Decision Run has been initialized

Decision RecommendCustomerProducts: Evaluate All Customer Products

Execute Rules <EvaluateOneProduct> for a collection Products of the type Product