As I start to use dataiku to standardize analysis chains, there are product specific data preparation nodes that are not always needed. Currently, if I want multiple products to use the same flow, I have to embed custom logic in many recipe nodes. It would be great, if nodes could be bypassed (grayed out) but not taken from the flow, using some global variable.
Couldn't you do this by building specific Scenarios that only run the required parts of your flow? Note that a Scenario can run another Scenario so you can easily nest/group your flow using this technique.
@Turribeach Yes I could do that but that is pretty complicated as well. The business logic could be 10+ nodes in the flow. I imagine a base flow that works out of the box if a product doesn't have dirty data. If a product does add in nodes to massage their data, they are tagged with a product bypass flag so the scenario can just build the base nodes, which would hit the product-specific nodes if they exist and are not bypassed.
Thanks for the feedback @info-rchitect , we've logged this request and will let you know if we have any updates.
Only members of the Community can comment.