About edited clustering label

Options
aya
aya Partner, Dataiku DSS Core Designer, Dataiku DSS & SQL, Dataiku DSS ML Practitioner, Dataiku DSS Core Concepts, Dataiku DSS Adv Designer, Registered Posts: 18 Partner

Hi,

Is there a way to reflect the clustering label to output dataset?
If I select "Create a recipe to perform clustering" in "Deploy a clustering model", the edited clustering label will not be reflected to output dataset. Renaming "cluseter_0" or "cluster_1" in the results window still leave "cluseter_0" or "cluster_1" in the output dataset.

If I select "Deploy a retrainable model to flow" and apply recipe, the edited clustering label will reflected to output dataset. But, I won't use this method in this case because of more step than left method.

Best Regard,
Aya

Tagged:

Best Answer

  • MarcH
    MarcH Dataiker Posts: 26 Dataiker
    Answer ✓
    Options

    Hi,

    At the moment, the clustering recipe does not take into account the edited cluster labels from the original model. I've added an issue to our backlog regarding this behavior.

    In the meantime, the workaround would indeed be to deploy a retrainable model, as you mentioned.

Setup Info
    Tags
      Help me…