Olivier Labayle

Follow up PR: https://github.com/JuliaAI/MLJBase.jl/pull/767

Same thing here, a simple loop with only an SVM in the Stack produces the error on my side if that helps: ```julia metalearner = EpsilonSVR() models = (model=EpsilonSVR(),) mystack...

Yes I agree that the current design leaves no other choice that of hyperparameter or global variable, both of them are a bit unatural since the caching requirement has already...

> Perhaps the resampling update to `Stack` is responsible. @ablaom Indeed, I've had a quick look. It seems the problem is coming from a source node operation (unfortunately this is...

I think this is be coming from the function `selectrows` which has to store the data because it is a partial function. In that sense I don't think anything is...

> Ah-haaa! Yes that makes sense. > > @olivierlabayle Thanks for taking time to diagnose this problem. This is an interesting case I had never thought of: data being stored...

Thank you for adding the support for propagating missing values! I think I have identified a bug if the **first** value in a vector is missing: ```julia using MLJModels, CategoricalArrays,...

I can give it a try if it's as easy as my suggestion, can you grant me access to the repo?

https://github.com/JuliaAI/MLJModels.jl/pull/468

I understand, thank you both for the clarification! Maybe the sentence that would benefit from clarification is the following: "If the core algorithm being wrapped requires data in a different...