[mlpack-git] [mlpack/mlpack] Universal B tree implementation (#746)

[Ryan Curtin]

> +template<typename BoundType, typename MatType>
> +bool UBTreeSplit<BoundType, MatType>::SplitNode(BoundType& bound,
> +                                                MatType& data,
> +                                                const size_t begin,
> +                                                const size_t count,
> +                                                SplitInfo& splitInfo)
> +{
> +  constexpr size_t order = sizeof(AddressElemType) * CHAR_BIT;
> +  if (begin == 0 && count == data.n_cols)
> +  {
> +    // Calculate all addresses.
> +    InitializeAddresses(data);
> +
> +    // Probably this is not a good idea. Maybe it is better to get
> +    // a number of distinct samples and find the median.
> +    std::sort(addresses.begin(), addresses.end(), ComparePair);

I think this is an okay approach for now.  Generally tree construction time is pretty minimal compared to the actual search time, so I don't know if there's much need to actually do the sampling to calculate the median.

-- 
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/746/files/f17843fc8e8ef2b1c4b04d572c521575f20a1f3c#r76270969
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://mailman.cc.gatech.edu/pipermail/mlpack-git/attachments/20160825/bc3ffee0/attachment.html>