[mlpack-git] [mlpack/mlpack] Hilbert R tree (#664)

[lozhnikov]

I fixed those issues. And I did the refactoring of all RectangleTree classes. Now they should support `localDataset`. I had to rewrite DiscreteHilbertValue entirely.
But I am not sure of `dataset`. As stated in `src/mlpack/core/tree/example_tree.hpp` each tree should satisfy that definition. Particulary `size_t Point(const size_t i) const` returns the index of a particular point of the node. Maybe we should not remove `dataset` from the RectangleTree class.

RecursiveHilbertValue is slow since there are a lot of levels of the recursion process. Indeed each double value belongs to \[-2^300, 2^300\] (roughly). I think we can use RecursiveHilbertValue in order to test DiscreteHilbertValue (they should give equal results).

I have not tested HilbertValueType separately yet. I have some ideas. I'll try them tomorrow.

---
You are receiving this because you are subscribed to this thread.
Reply to this email directly or view it on GitHub:
https://github.com/mlpack/mlpack/pull/664#issuecomment-223785541
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://mailman.cc.gatech.edu/pipermail/mlpack-git/attachments/20160604/d462d8fc/attachment.html>