kdt貌似是根号的...（虽然我没分析过复杂度
而且根据以往的经验，似乎非曼哈顿距离的情况下方差划分比直接轮流划分效果更差QAQ
by qq872191552 @ 2017-07-15 21:45:42
你三维K-D树复杂度对吗？
by zcysky @ 2017-07-16 11:53:11
@[zcysky](/space/show?uid=2978)
或许是一维排序然后KD-tree...
by qq872191552 @ 2017-07-16 14:17:04
@[zcysky](/space/show?uid=2978) 我是一维sort+二维kdtree，要不然该全T啊
by Rockdu @ 2017-07-16 19:16:06
已经用cdq解决了，谢谢！
by Rockdu @ 2017-07-20 11:30:29
难道不是sort + BIT + CDQ？
by KKarshilov @ 2017-11-18 22:00:29
@[Rockdu](/space/show?uid=39907)
佬惊现
by MagentaSS @ 2018-11-03 15:57:27