Matthieu Gomez

Hey! I need to look at this more closely. That being said, i would really like to avoid adding two new dependences.

Okay, I am back from holidays. Before I look into it, - I think it would be useful to remove the two new dependences, so that I can look at...

For your cross_difference function, you may be right. Could you explain a bit more what you want to change (and why)?

Isn't `AchdouHanLasryLionsMoll_OneAsset.jl` enough? What would you like to have in `AiyagariContinuousTime`?

No, sorry, I don’t have the time to do it.

Maybe it would be more elegant to restrict it to functions for which all arguments are AbstractVector, i.e. `@assert all(x -> x isa AbstractVector, xs)`. AFAIU, this does not hurt...

Another thing is that `spreadmissing` always returns a `Vector`, but some functions may return `AbstractVectors` that are not `Vectors` (e.g. . `cut` returns a `CategoricalVector`) So it may be better...

It'd be great to have such a function! It's just hard to find the right abstraction, so it's good to experiment.

Parallelization is great. The functions I optimize generally use a lot of temporary memory, so I generally add a set preallocated arrays as an argument. When optimizing the function, I...

Great. Thanks for the explanation.