The endemic difficulty of manipulating array on Dataiku

Hello,

I take time of these few moments of respite to write a post on this subject that many of my collaborators meet and whihch I have already confronted this problem with Dataiku's staff but which have only led to complex or very restrictive palliative solutions and very unfriendly for no-code user :

- Managing array variable.

It's a common type of variable in a lot of use case but when you are confronted to use this type in DSS it could be quickly tought to deal with it.

Let me explain :

- Much of formula "array" function never worked except for very basics things like "len" or get a sub-array element. (confirmed by my last point with DSS Tier2 Support)

- Most of the time : parsing array in a step doesnt work you > have to build one more step (for exemple, in python notebook ) for loads it in JSON or manipulate array's variable for any project.

- Most of time : when you have this type of variable as storage type in your ouput/input dataset with a meaning as a array type too, you are pretty sure every "no-code" step/recipe iterate on it in background as a string variable.

>screenshot is not representative but it is and example for this post