[mlpack-git] [mlpack/mlpack] Vantage point tree (#708)

[Ryan Curtin]

> +  }
> +}
> +
> +template<typename MetricType,
> +         typename StatisticType,
> +         typename MatType,
> +         template<typename BoundMetricType, typename...> class BoundType,
> +         template<typename SplitBoundType, typename SplitMatType>
> +             class SplitType>
> +template<typename RuleType>
> +void VantagePointTree<MetricType, StatisticType, MatType, BoundType, SplitType>::
> +DualTreeTraverser<RuleType>::Traverse(
> +    const size_t queryIndex,
> +    VantagePointTree<MetricType, StatisticType, MatType, BoundType, SplitType>&
> +        referenceNode)
> +{

I believe you are right that `traversalInfo` is not modified by a single-tree traversal.

I don't think it's necessary to add a `SingleTreeTraverser` object to `DualTreeTraverser`, but instead simply implement both in a single class:

```
template<typename RuleType>
class Traverser
{
  void Traverse(const size_t queryIndex, TreeType& referenceNode);
  void Traverse(TreeType& queryNode, TreeType& referenceNode); 
}
```

Then inside the `VantagePointTree` class you could define this class as both the single-tree and dual-tree traverser type.  I don't think there would be an issue with `numBaseCases` or `numPrunes` or `numScores` that way (or whichever of those statistics are tracked).  Maybe I have overlooked something here?

---
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/708/files/300882ac96e7a663e3e303ca0c45c14c6fafe1a6#r71911629
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://mailman.cc.gatech.edu/pipermail/mlpack-git/attachments/20160722/173023e2/attachment.html>