This is an extension for the ELKI Data Mining Toolkit to use the PH-tree index.
Build the package using Maven mvn package
, then add the resulting
target/elki-phtree-1.0.jar
to your ELKI class path.
To use the ph-tree, set the -db.index
parameter to
tree.spatial.ph.MemoryPHTree
and use an algorithm using
Euclidean distance (other Minkowski norms and squared Euclidean distance are
also supported), and either radius or k-nearest-neighbor queries.
The PH-tree is a multi-dimensional indexing and storage structure. By default it stores k-dimensional keys (points) consisting of k 64bit-integers. However, it can also be used to efficiently store floating point values or k-dimensional rectangles. It supports kNN queries, range queries, window queries and fast update/moving of individual entries.
The PH-tree was developed at ETH Zurich and first published in: "The PH-Tree: A Space-Efficient Storage Structure and Multi-Dimensional Index" (PDF), Tilmann Zäschke, Christoph Zimmerli and Moira C. Norrie, Proceedings of Intl. Conf. on Management of Data (SIGMOD), 2014
The current version of the PH-tree is discussed in more detail here: (PDF) (2015).
Contact: {zaeschke,zimmerli,norrie)@inf.ethz.ch