wrong answer

[APernatii]

So, Im not sure what is happening, but I try to use the following stuff

using Neighborhood
using Distances
queryPoint = [2.0,1.0]
dataX = [0.5 0.5; 1.5 0.5]
tree = Neighborhood.searchstructure(KDTree,dataX,Euclidean(),reorder = false)
idx,ds = Neighborhood.knn(tree, queryPoint, nonearpoints)

And I got the answer
idx,ds = Neighborhood.knn(tree, queryPoint, nonearpoints)
([1, 2], [1.5811388300841898, 1.5811388300841898])

that is clearly whong, because if I check the distance
Euclidean()(queryPoint ,dataX[2,:])
I got the answer: 0.7071067811865476

What am I doing wrong here? Please help

[Datseris]

can you run this again without the reoder = false keyword?

[APernatii]

same answer

[Datseris]

in your original code nonearpoints is undefined.

[APernatii]

Sorry
nonearpoints = 2

[Datseris]

Unfortunately I don't think I can help you because this doesn't appear to be an interface problem (Neighborhoods.jl only provodes the interface). What I would recomend is that you reproduce this using NearestNeighbors.jl directly, and open an issue there. The developers will have better chances understanding what is wrong.

[APernatii]

Oh, I see, Thanks a lot!