I want to test a learner (eg SOTA-learner) with different parameter options of the node (eg. in the SOTA-case the distance metric can have two values). How can I make a workflow that chooses each of these options?
You should pass a parameter as a flow variable using a loop. Please find a sample workflow which clusters the data with different numbers of clusters (different number of clusters is passed for each loop iteration, the number of clusters is regulated by a flow variable in the k-Means node).
Thanks for your reply Anna.
I know the mechanism of using flow variables and that works well.
But I think there is a misunderstanding (maybe I was not exactly clear), my question runs a bit deeper: how can I query a node for the elements that have an option list and after that how can I run programmatically over the different options (ie SOTA-learner could have in the future maybe 3 distance metrics instead of the 2 now, And maybe there comes another optionlist. I would like to have a workflow that incorporates that without manual intervention.).