Daniel Karrasch

This seems to close #65.

You can find the `PeriodicEuclidean` distance over at [Distances.jl](https://github.com/JuliaStats/Distances.jl). And then you can pass it as the metric argument to the tree constructor here. You'll need to use a BallTree,...

You could use `StaticArrays` ```julia using StaticArrays, NearestNeighbors data = rand(3, 10^4); k = 3; kdtree = KDTree(data); points = [@SVector rand(3) for i in 1:10] idxs, dists = knn(kdtree,...

@KristofferC may help out if I'm saying something wrong, as I haven't written or modified any parts of the code, just scanned quickly through it some time ago. If I...

Since it's very early in the v1.10 release cycle, we have plenty of time, so I'll release patches for the fundamental linalg packages that would pick up the changes and...

@nanosoldier `runtests()`

Currently probably not. But what should a `view` on a function-based `AbstractLinear` object be? Probably any operation downstream will treat it like an `AbstractArray` and start to get the entries...

From the nanosoldier run, here is another common pattern of failure with this PR: ```julia MethodError: no method matching lmul!(::LinearAlgebra.QRCompactWYQ{Float64, Matrix{Float64}, Matrix{Float64}}, ::SparseArrays.SparseMatrixCSC{Float64, Int64}) ERROR: LoadError: MethodError: no method matching...

I see. Thanks for clarifying. It would be a very strong feature enhancement if the piracy check could, for union arguments as above, something along the lines "what if I...

First I was afraid this would bring back #123 (see #124 for the offending method signature), but this works since you keep the method signature as is. Unfortunately, on my...