Kyle Daruwalla
Kyle Daruwalla
This is great, thanks for the contribution! It might make more sense to add a new top-level section for "Tutorials" instead of adding more to the first page? Also, MLUtils.jl...
MLDataPattern.jl had time series functions, but they were tricky to work with and didn't compose as well with the rest of the package. In general, time dimensions are hard to...
I agree with Brian that we should look into reducing the load time contributed by Floops first. It would be nicer as a user to not have to load different...
I share Lorenz's view here about waiting to see what people say. Also, keeping it in means we can release now, so I suggest doing that and moving this discussion...
Yes, we should eventually migrate the docs and add more tutorials about typical use cases.
As a temporary workaround, you can do ```julia batches = [[1, 2], [3]] MLUtils.batch(rpad.(batches, 2, 0)) ``` Probably a variation of this where we pad as we iterate `batches` would...
Closing, since I guess the solution here is to pass `f` as a named tuple to `mapobs`?
The current `mapobs` story is very confusing. `mapobs(fs::NamedTuple, data)` will call a separate `f in fs` on each observation of data. Why does `NamedTupleData` exist? Can't it just be `(;...
Maybe something like this would work: ```julia Base.getindex(data::MLUtils.MappedData, idx::Integer) = data.f(getobs(data.data, idx)) Base.getindex(data::MLUtils.MappedData, idxs::AbstractVector) = batch(map(Base.Fix1(getindex, data), idxs)) ```
But if your container is a (named) tuple, why not just do `mapobs(f, data[2])`?