apaam/mlpack
1
0
Fork 0
Code Issues Pull Requests Packages Projects Releases Wiki Activity

Some fairly serious refactoring, but it works.

Browse Source 
There are significant optimizations to still be made, but at this point we get
some acceleration over the naive algorithm.
This commit is contained in:
Ryan Curtin
2015-03-12 15:42:27 -04:00
parent 44fa159677
commit 56cbc1c3bb
4 changed files with 37 additions and 130 deletions
Download Patch File Download Diff File Expand all files Collapse all files
src/mlpack/methods/kmeans/dual_tree_kmeans_impl.hpp
+3 -2
View File
@@ -88,6 +88,7 @@ double DualTreeKMeans<MetricType, MatType, TreeType>::Iterate(
typename TreeType::template BreadthFirstDualTreeTraverser<RulesType>
traverser(rules);
tree->Stat().ClustersPruned() = 0; // The constructor sets this to -1.
traverser.Traverse(*centroidTree, *tree);
distanceCalculations += rules.DistanceCalculations();
@@ -220,8 +221,8 @@ void DualTreeKMeans<MetricType, MatType, TreeType>::TreeUpdate(
// We have to set the closest query node to NULL because the cluster tree will
// be rebuilt.
node->Stat().ClosestQueryNode() = NULL;
// node->Stat().MaxQueryNodeDistance() = DBL_MAX;
// node->Stat().MinQueryNodeDistance() = DBL_MAX;
node->Stat().MaxQueryNodeDistance() = DBL_MAX;
node->Stat().MinQueryNodeDistance() = DBL_MAX;
for (size_t i = 0; i < node->NumChildren(); ++i)
TreeUpdate(&node->Child(i), clusters, clusterDistances);
src/mlpack/methods/kmeans/dual_tree_kmeans_rules.hpp
-2
View File
@@ -68,8 +68,6 @@ class DualTreeKMeansRules
TraversalInfoType traversalInfo;
double IterationUpdate(TreeType& referenceNode);
bool IsDescendantOf(const TreeType& potentialParent, const TreeType&
potentialChild) const;
src/mlpack/methods/kmeans/dual_tree_kmeans_rules_impl.hpp
+33 -125
View File
@@ -50,13 +50,9 @@ inline force_inline double DualTreeKMeansRules<MetricType, TreeType>::BaseCase(
const size_t queryIndex,
const size_t referenceIndex)
{
// Log::Info << "Base case, query " << queryIndex << " (" << mappings[queryIndex]
// << "), reference " << referenceIndex << ".\n";
// Collect the number of clusters that have been pruned during the traversal.
// The ternary operator may not be necessary.
const size_t traversalPruned = (traversalInfo.LastReferenceNode() != NULL) ?
// traversalInfo.LastReferenceNode()->Stat().Iteration() == iteration) ?
traversalInfo.LastReferenceNode()->Stat().ClustersPruned() : 0;
// It's possible that the reference node has been pruned before we got to the
@@ -87,8 +83,6 @@ inline force_inline double DualTreeKMeansRules<MetricType, TreeType>::BaseCase(
if (visited[referenceIndex] + traversalPruned == centroids.n_cols)
{
// Log::Warn << "Commit reference index " << referenceIndex << " to cluster "
// << assignments[referenceIndex] << ".\n";
newCentroids.col(assignments[referenceIndex]) +=
dataset.col(referenceIndex);
++counts(assignments[referenceIndex]);
@@ -99,12 +93,9 @@ inline force_inline double DualTreeKMeansRules<MetricType, TreeType>::BaseCase(
template<typename MetricType, typename TreeType>
double DualTreeKMeansRules<MetricType, TreeType>::Score(
const size_t /* queryIndex */,
const size_t queryIndex,
TreeType& referenceNode)
{
// Update from previous iteration, if necessary.
// IterationUpdate(referenceNode);
// No pruning here, for now.
return 0.0;
}
@@ -114,6 +105,7 @@ double DualTreeKMeansRules<MetricType, TreeType>::Score(
TreeType& queryNode,
TreeType& referenceNode)
{
// This won't happen with the root since it is explicitly set to 0.
if (referenceNode.Stat().ClustersPruned() == size_t(-1))
referenceNode.Stat().ClustersPruned() =
referenceNode.Parent()->Stat().ClustersPruned();
@@ -124,35 +116,34 @@ double DualTreeKMeansRules<MetricType, TreeType>::Score(
// We also have to update things if the closest query node is null. This can
// probably be improved.
if (score != DBL_MAX || referenceNode.Stat().ClosestQueryNode() == NULL)
const double minDistance = referenceNode.MinDistance(&queryNode);
const double maxDistance = referenceNode.MaxDistance(&queryNode);
distanceCalculations += 2;
score = PellegMooreScore(queryNode, referenceNode, minDistance);
if (referenceNode.Stat().MaxQueryNodeDistance() == DBL_MAX &&
referenceNode.Parent() != NULL &&
referenceNode.Parent()->Stat().MaxQueryNodeDistance() != DBL_MAX)
{
// Can we update the minimum query node distance for this reference node?
const double minDistance = referenceNode.MinDistance(&queryNode);
const double maxDistance = referenceNode.MaxDistance(&queryNode);
distanceCalculations += 2;
if (maxDistance < referenceNode.Stat().MaxQueryNodeDistance())
{
referenceNode.Stat().ClosestQueryNode() = (void*) &queryNode;
referenceNode.Stat().MinQueryNodeDistance() = minDistance;
referenceNode.Stat().MaxQueryNodeDistance() = maxDistance;
// referenceNode.MaxDistance(&queryNode);
// ++distanceCalculations;
return 0.0; // Pruning is not possible.
}
referenceNode.Stat().ClosestQueryNode() =
referenceNode.Parent()->Stat().ClosestQueryNode();
referenceNode.Stat().MaxQueryNodeDistance() =
referenceNode.Parent()->Stat().MaxQueryNodeDistance();
}
else if (IsDescendantOf(
*((TreeType*) referenceNode.Stat().ClosestQueryNode()), queryNode))
{
// Just update.
referenceNode.Stat().ClosestQueryNode() = (void*) &queryNode;
referenceNode.Stat().MinQueryNodeDistance() = minDistance;
referenceNode.Stat().MaxQueryNodeDistance() =
referenceNode.MaxDistance(&queryNode);
++distanceCalculations;
return 0.0; // Pruning is not possible.
}
score = PellegMooreScore(queryNode, referenceNode, minDistance);
if (maxDistance < referenceNode.Stat().MaxQueryNodeDistance() ||
referenceNode.Stat().ClosestQueryNode() == NULL)
{
referenceNode.Stat().ClosestQueryNode() = (void*) &queryNode;
referenceNode.Stat().MinQueryNodeDistance() = minDistance;
referenceNode.Stat().MaxQueryNodeDistance() = maxDistance;
}
else if (IsDescendantOf(*((TreeType*)
referenceNode.Stat().ClosestQueryNode()), queryNode))
{
referenceNode.Stat().ClosestQueryNode() == (void*) &queryNode;
referenceNode.Stat().MinQueryNodeDistance() = minDistance;
referenceNode.Stat().MaxQueryNodeDistance() = maxDistance;
}
if (score == DBL_MAX)
@@ -176,17 +167,16 @@ double DualTreeKMeansRules<MetricType, TreeType>::Score(
else if (referenceNode.Stat().ClustersPruned() +
visited[referenceNode.Descendant(0)] == centroids.n_cols)
{
for (size_t i = 0; i < referenceNode.NumPoints(); ++i)
for (size_t i = 0; i < referenceNode.NumDescendants(); ++i)
{
const size_t cluster = assignments[referenceNode.Point(i)];
newCentroids.col(cluster) += dataset.col(referenceNode.Point(i));
const size_t cluster = assignments[referenceNode.Descendant(i)];
newCentroids.col(cluster) += dataset.col(referenceNode.Descendant(i));
counts(cluster)++;
}
}
}
return score;
// return 0.0;
}
template<typename MetricType, typename TreeType>
@@ -205,87 +195,6 @@ double DualTreeKMeansRules<MetricType, TreeType>::Rescore(
const double oldScore) const
{
return oldScore;
// if (oldScore == DBL_MAX)
// return oldScore; // We can't unprune something. This shouldn't happen.
// return ElkanTypeScore(queryNode, referenceNode, oldScore);
}
template<typename MetricType, typename TreeType>
inline double DualTreeKMeansRules<MetricType, TreeType>::IterationUpdate(
TreeType& referenceNode)
{
Log::Fatal << "Update! Why!\n";
if (referenceNode.Stat().Iteration() == iteration)
return 0;
const size_t itDiff = iteration - referenceNode.Stat().Iteration();
referenceNode.Stat().Iteration() = iteration;
referenceNode.Stat().ClustersPruned() = (referenceNode.Parent() == NULL) ?
0 : referenceNode.Parent()->Stat().ClustersPruned();
referenceNode.Stat().ClosestQueryNode() = (referenceNode.Parent() == NULL) ?
NULL : referenceNode.Parent()->Stat().ClosestQueryNode();
if (referenceNode.Stat().ClosestQueryNode() != NULL)
{
referenceNode.Stat().MinQueryNodeDistance() =
referenceNode.MinDistance((TreeType*)
referenceNode.Stat().ClosestQueryNode());
referenceNode.Stat().MaxQueryNodeDistance() =
referenceNode.MaxDistance((TreeType*)
referenceNode.Stat().ClosestQueryNode());
distanceCalculations += 2;
}
if (itDiff > 1)
{
// referenceNode.Stat().BestMaxDistance() = DBL_MAX;
referenceNode.Stat().MinQueryNodeDistance() = DBL_MAX;
referenceNode.Stat().MaxQueryNodeDistance() = DBL_MAX;
}
else
{
if (referenceNode.Stat().MinQueryNodeDistance() != DBL_MAX)
{
// Update the distance to the closest query node. If this node has an
// owner, we know how far to increase the bound. Otherwise, increase it
// by the furthest amount that any centroid moved.
if (referenceNode.Stat().Owner() < centroids.n_cols)
{
referenceNode.Stat().MinQueryNodeDistance() +=
clusterDistances(referenceNode.Stat().Owner());
referenceNode.Stat().MaxQueryNodeDistance() +=
clusterDistances(referenceNode.Stat().Owner());
}
else
{
referenceNode.Stat().MinQueryNodeDistance() +=
clusterDistances(centroids.n_cols);
referenceNode.Stat().MaxQueryNodeDistance() +=
clusterDistances(centroids.n_cols);
}
}
else
{
referenceNode.Stat().MinQueryNodeDistance() = DBL_MAX;
referenceNode.Stat().MaxQueryNodeDistance() = DBL_MAX;
}
}
// if (referenceNode.Stat().BestMaxDistance() != DBL_MAX)
// {
// if (referenceNode.Stat().Owner() < centroids.n_cols)
// referenceNode.Stat().BestMaxDistance() +=
// clusterDistances(referenceNode.Stat().Owner());
// else
// referenceNode.Stat().BestMaxDistance() +=
// clusterDistances(centroids.n_cols);
// }
// }
return 1;
}
template<typename MetricType, typename TreeType>
@@ -308,12 +217,11 @@ double DualTreeKMeansRules<MetricType, TreeType>::ElkanTypeScore(
{
// We have to calculate the minimum distance between the query node and the
// reference node's best query node. First, try to use the cached distance.
// const double minQueryDistance = queryNode.Stat().FirstBound();
const double minQueryDistance = queryNode.Stat().FirstBound();
if (queryNode.NumDescendants() == 1)
{
const double score = ElkanTypeScore(queryNode, referenceNode,
interclusterDistances[queryNode.Descendant(0)]);
// Log::Warn << "Elkan scoring: " << score << ".\n";
return score;
}
else
@@ -344,7 +252,7 @@ double DualTreeKMeansRules<MetricType, TreeType>::ElkanTypeScore(
template<typename MetricType, typename TreeType>
double DualTreeKMeansRules<MetricType, TreeType>::PellegMooreScore(
TreeType& /* queryNode */,
TreeType& queryNode,
TreeType& referenceNode,
const double minDistance) const
{
src/mlpack/methods/kmeans/dual_tree_kmeans_statistic.hpp
+1 -1
View File
@@ -20,7 +20,7 @@ class DualTreeKMeansStatistic
closestQueryNode(NULL),
minQueryNodeDistance(DBL_MAX),
maxQueryNodeDistance(DBL_MAX),
clustersPruned(0),
clustersPruned(size_t(-1)),
iteration(size_t() - 1),
firstBound(DBL_MAX),
secondBound(DBL_MAX),