distance computations in L3 nearest neighbor

[esrasuel]

I think there is a problem with the computation of distances in L3 nearest neighbor function due to a mix up of lat/lng order.

For example, if you look at the output 7 in nearest-neighbor-faster.ipynb :

the distance between POINT (24.85584 60.20727) and POINT (24.85745 60.20749) is 180.521584

when I try to replicate with haversine or sklearn , I get 92 meters instead.

I also tried switching lat/lng pairs, then you get the 180.521584 meter output. But my understanding is that the correct distance should be 92m.

[HTenkanen]

@esrasuel Thanks for pointing this out! We were also approached by another user who pointed out the same thing. We will fix this.