Can you choose the order that items run?

FarSideFeb
Level 3
Can you choose the order that items run?

My company started utilizing Dataiku as a BI tool. Before deciding to completely immerse myself into Dataiku, I played around to see if Dataiku could do specific things that I had been doing in a different BI tool (we will call this creating flow zones 2 and 3). Once I realized Dataiku could do what I wanted, I started adding to what I had played around with by building what I would label as flow zone 1, as this is how I would have started building my workflow if I had known from the start that Dataiku could do everything I wanted it to. All of this to say, can I tell Dataiku to run the flow zones in a specific order or do I have to start from scratch and rebuild everything?


Operating system used: Windows

0 Kudos
3 Replies
AlexT
Dataiker

Hi @FarSideFeb,

Flow Zone won't impact the building of datasets in your flow, you can re-arrange and move items to different Flow zones at any time but can't build specific flowzones. 

To build your Flow you can use Recursive Builds, select the right-most dataset/folders or specific dataset in your from your and it will build all dependencies as explained here:

https://knowledge.dataiku.com/latest/courses/flow-views-and-actions/building-strategies-concept-summ... 

You can automate your Flow build by adding the right most datasets to a scenario with build dataset/folder steps. 

Let me know if that helps or if you have any other questions. 

 

0 Kudos
FarSideFeb
Level 3
Author

Good morning @AlexT , thank you for your response!

I should have worded my question a little differently. The items within "flow zone 2/3" were built first, then I built the items within "flow zone 1". If I had done this same thing in Alteryx, the items within "flow zone 2/3" would run first, then the items within "flow zone 1" would run second. I need the query in flow zone 1 to run first, as the data from flow zone 1, is used to run the query in flow zone 2/3. 

0 Kudos
AlexT
Dataiker

Hi,

If I understand what your flow looks like then using a scenario is likely the best option.

Add the output datasets you want to build from Flow Zone 1 to be built steps then The output dataset from Flow Zone 2 etc. 

This will also depend on how your Flow Zones are connected if they are independent and no dependencies in another flow zone vs if the output dataset from FlowZone3 is within the same flow as FlowZone1. In this case, you can select output dataset/s from FlowZone3 with recursive build and the flow will first build FlowZone1..2 then eventually 3 from Left -> Right. 

Let me know if that makes sense. 

Thanks,