You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
In particular, FindNearestNeighbors() function has been completely rewritten, and now 1) actually works correctly, 2) uses Euclidean metric, 3) has more user friendly calling sequence, 4) is recursive and more readable (as fast as before). Some new functions had to be added to TKDTree for this, such as Distance(), which finds distance between a random point and a point in the kdtree, and DistanceToNode() which finds the distance between a point and a node of the kdtree, GetNodePointsIndexes() which returns ranges of indexes of original points, contained in a given node. Also, the test macro has been updated with a test function, that compares TKDTree::FindNearestNeighbors() results with simple sorting.
fixed also some problem in the TMath test
git-svn-id: http://root.cern.ch/svn/root/trunk@26356 27541ba8-7e3a-0410-8455-c3a389f83636
0 commit comments