david-vicente

> One of the motivations of `with_lengthscale` is that it provides a unified API for `ScaleTransform` and `ARDTransform` (although one can save allocations by using `ARDTransform` directly). I thought that...

Sorry for the trouble. At the time I was having a hard time using the existing methods because I wanted to perform AD tests for a kernel with tuples of...

Why isn't there a period length parameter as in [gpytorch](https://docs.gpytorch.ai/en/stable/kernels.html#periodickernel)?

Thank you @st--. I did look to the `PeriodicTransform` at the time, but it wasn't obvious how to use it to achieve equivalent kernels from the documentation. Maybe we should...

Can we do the same for multi-dimensional inputs?

I don't like abbreviations used in programming so I'm in favor of removing the `MO` from the concrete types' names and using `MultiOutput` whenever there is ambiguity. Off-topic: Is it...

I like the isotopic prefix, it has a clear meaning. The first time I read the documentation and the code for `MOinputs` it took me quite a while to understand...

> Copied from Slack: [Offtopic] Is this Julia's Slack, or a private one?

I also noticed that the signatures of the constructors for `MOARGP`, `MOSVGP` an `MOVGP` require that the input labels variable `y` is a `AbstractVector{

Trying your suggestion gives me: ```Julia BoundsError: attempt to access 1-element Array{Int64,1} at index [2] Stacktrace: [1] getindex at ./array.jl:809 [inlined] [2] #76 at /home/myuser/.julia/packages/AugmentedGaussianProcesses/93OMX/src/inference/analyticVI.jl:49 [inlined] [3] ntuple(::AugmentedGaussianProcesses.var"#76#77"{Float64,Array{Int64,1},Descent}, ::Int64) at...