Altair RISE

A program to recognize and reward our most engaged community members

Nominate Yourself Now!

Return only the last value from Loop

Hi There,
I started using rapidminer just today. There's one problem that I'm facing and I can't find the solution anywhere.
Objective: So I want to do clustering for n max iterations, for each n I want to see the centroid value, after this, I want to only work with the example set that the clustering algorithm returned in the last run.

I have done half of it, i.e, I have used Loop Parameter to see all n centroids. Each n is the output of the Loop, but the problem here is that, each n also returns a corresponding example set, so I end up with n example sets which I do NOT want.
I only want the final example set. How can I get that example set?

Regards

Find more posts tagged with

AI Studio

Clustering

Loops + Branches

k-Means Clustering

Accepted answers

lionelderkrikor

@Ray8989,

Can you please guide me where and how to use macro?

Do you refer to Brian's method ?
If yes, you will need the Extract Macro from Collection operator. This operator will detect the size of your collection (equal to n in your case) and store this value as a macro. Then you can use this macro inside the Select operator to "capture" the last example set of your collection as described by Brian.
To use Extract Macro from Collection operator you will need to install the Jackhammer extension (this extension must be installed from the MarketPlace).

Please check the process in attached file and tell me if it answers to your need.

Regards,

Lionel

Loop_parameters_clustering_2.rmp

All comments

Telcontar120

If you have a collection of examplesets being returned, you can just use the "Select" operator to specify the one you want to keep (you could even do this automatically by making a macro for the loop iterations and using that as the number of the value to return in the Select operator). If you are trying to do something more complex perhaps you can post your xml so it is a bit clearer.

lionelderkrikor

Hi @Ray8989,

Maybe I have misunderstood your final goal,
but you can apply the clustering model with max_runs = n to your dataset in parallel of the Loop Parameters operator.
Please look at the process in attached file to understand what I mean...

Regards,

Lionel

Loop_parameters_clustering.rmp

Ray8989

Thanks

Ray8989

lionelderkrikor, It does work. And that is what the end goal was. But I want to use only one instance of Clustering.
Can you please guide me where and how to use macro?

lionelderkrikor

@Ray8989,

Can you please guide me where and how to use macro?

Loop_parameters_clustering_2.rmp

Ray8989

@lionelderkrikor works exactly as I wanted! Thank you!